2010-09-20 22:08:53 +02:00
|
|
|
<!-- doc/src/sgml/protocol.sgml -->
|
2001-06-23 01:27:48 +02:00
|
|
|
|
|
|
|
<chapter id="protocol">
|
|
|
|
<title>Frontend/Backend Protocol</title>
|
|
|
|
|
2008-01-09 06:27:22 +01:00
|
|
|
<indexterm zone="protocol">
|
|
|
|
<primary>protocol</primary>
|
|
|
|
<secondary>frontend-backend</secondary>
|
|
|
|
</indexterm>
|
|
|
|
|
2001-06-23 01:27:48 +02:00
|
|
|
<para>
|
2003-09-11 23:42:20 +02:00
|
|
|
<productname>PostgreSQL</productname> uses a message-based protocol
|
2003-04-16 00:51:18 +02:00
|
|
|
for communication between frontends and backends (clients and servers).
|
|
|
|
The protocol is supported over <acronym>TCP/IP</acronym> and also over
|
|
|
|
Unix-domain sockets. Port number 5432 has been registered with IANA as
|
|
|
|
the customary TCP port number for servers supporting this protocol, but
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
in practice any non-privileged port number can be used.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
This document describes version 3.0 of the protocol, implemented in
|
2003-09-11 23:42:20 +02:00
|
|
|
<productname>PostgreSQL</productname> 7.4 and later. For descriptions
|
2003-04-16 00:51:18 +02:00
|
|
|
of the earlier protocol versions, see previous releases of the
|
|
|
|
<productname>PostgreSQL</productname> documentation. A single server
|
2017-11-21 19:56:24 +01:00
|
|
|
can support multiple protocol versions. The initial startup-request
|
|
|
|
message tells the server which protocol version the client is attempting to
|
|
|
|
use. If the major version requested by the client is not supported by
|
|
|
|
the server, the connection will be rejected (for example, this would occur
|
|
|
|
if the client requested protocol version 4.0, which does not exist as of
|
|
|
|
this writing). If the minor version requested by the client is not
|
2020-09-01 00:33:37 +02:00
|
|
|
supported by the server (e.g., the client requests version 3.1, but the
|
2017-11-21 19:56:24 +01:00
|
|
|
server supports only 3.0), the server may either reject the connection or
|
|
|
|
may respond with a NegotiateProtocolVersion message containing the highest
|
|
|
|
minor protocol version which it supports. The client may then choose either
|
|
|
|
to continue with the connection using the specified protocol version or
|
|
|
|
to abort the connection.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
|
|
|
In order to serve multiple clients efficiently, the server launches
|
2017-10-09 03:44:17 +02:00
|
|
|
a new <quote>backend</quote> process for each client.
|
2003-04-16 00:51:18 +02:00
|
|
|
In the current implementation, a new child
|
|
|
|
process is created immediately after an incoming connection is detected.
|
|
|
|
This is transparent to the protocol, however. For purposes of the
|
2017-10-09 03:44:17 +02:00
|
|
|
protocol, the terms <quote>backend</quote> and <quote>server</quote> are
|
|
|
|
interchangeable; likewise <quote>frontend</quote> and <quote>client</quote>
|
2003-04-16 00:51:18 +02:00
|
|
|
are interchangeable.
|
|
|
|
</para>
|
|
|
|
|
2001-06-23 01:27:48 +02:00
|
|
|
<sect1 id="protocol-overview">
|
|
|
|
<title>Overview</title>
|
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The protocol has separate phases for startup and normal operation.
|
|
|
|
In the startup phase, the frontend opens a connection to the server
|
|
|
|
and authenticates itself to the satisfaction of the server. (This might
|
|
|
|
involve a single message, or multiple messages depending on the
|
|
|
|
authentication method being used.) If all goes well, the server then sends
|
|
|
|
status information to the frontend, and finally enters normal operation.
|
|
|
|
Except for the initial startup-request message, this part of the
|
|
|
|
protocol is driven by the server.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
During normal operation, the frontend sends queries and
|
|
|
|
other commands to the backend, and the backend sends back query results
|
2017-10-09 03:44:17 +02:00
|
|
|
and other responses. There are a few cases (such as <command>NOTIFY</command>)
|
2003-04-16 00:51:18 +02:00
|
|
|
wherein the
|
|
|
|
backend will send unsolicited messages, but for the most part this portion
|
|
|
|
of a session is driven by frontend requests.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Termination of the session is normally by frontend choice, but can be
|
|
|
|
forced by the backend in certain cases. In any case, when the backend
|
|
|
|
closes the connection, it will roll back any open (incomplete) transaction
|
|
|
|
before exiting.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Within normal operation, SQL commands can be executed through either of
|
2017-10-09 03:44:17 +02:00
|
|
|
two sub-protocols. In the <quote>simple query</quote> protocol, the frontend
|
2003-04-16 00:51:18 +02:00
|
|
|
just sends a textual query string, which is parsed and immediately
|
2017-10-09 03:44:17 +02:00
|
|
|
executed by the backend. In the <quote>extended query</quote> protocol,
|
2003-04-16 00:51:18 +02:00
|
|
|
processing of queries is separated into multiple steps: parsing,
|
|
|
|
binding of parameter values, and execution. This offers flexibility
|
|
|
|
and performance benefits, at the cost of extra complexity.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Normal operation has additional sub-protocols for special operations
|
2017-10-09 03:44:17 +02:00
|
|
|
such as <command>COPY</command>.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<sect2 id="protocol-message-concepts">
|
|
|
|
<title>Messaging Overview</title>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
All communication is through a stream of messages. The first byte of a
|
|
|
|
message identifies the message type, and the next four bytes specify the
|
|
|
|
length of the rest of the message (this length count includes itself, but
|
|
|
|
not the message-type byte). The remaining contents of the message are
|
|
|
|
determined by the message type. For historical reasons, the very first
|
|
|
|
message sent by the client (the startup message) has no initial
|
|
|
|
message-type byte.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
To avoid losing synchronization with the message stream, both servers and
|
|
|
|
clients typically read an entire message into a buffer (using the byte
|
|
|
|
count) before attempting to process its contents. This allows easy
|
|
|
|
recovery if an error is detected while processing the contents. In
|
|
|
|
extreme situations (such as not having enough memory to buffer the
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
message), the receiver can use the byte count to determine how much
|
2003-04-16 00:51:18 +02:00
|
|
|
input to skip before it resumes reading messages.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Conversely, both servers and clients must take care never to send an
|
|
|
|
incomplete message. This is commonly done by marshaling the entire message
|
|
|
|
in a buffer before beginning to send it. If a communications failure
|
|
|
|
occurs partway through sending or receiving a message, the only sensible
|
|
|
|
response is to abandon the connection, since there is little hope of
|
|
|
|
recovering message-boundary synchronization.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
2003-04-16 00:51:18 +02:00
|
|
|
</sect2>
|
|
|
|
|
|
|
|
<sect2 id="protocol-query-concepts">
|
|
|
|
<title>Extended Query Overview</title>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
In the extended-query protocol, execution of SQL commands is divided
|
|
|
|
into multiple steps. The state retained between steps is represented
|
2017-10-09 03:44:17 +02:00
|
|
|
by two types of objects: <firstterm>prepared statements</firstterm> and
|
|
|
|
<firstterm>portals</firstterm>. A prepared statement represents the result of
|
2011-09-16 06:42:53 +02:00
|
|
|
parsing and semantic analysis of a textual query string.
|
|
|
|
A prepared statement is not in itself ready to execute, because it might
|
2017-10-09 03:44:17 +02:00
|
|
|
lack specific values for <firstterm>parameters</firstterm>. A portal represents
|
2003-04-16 00:51:18 +02:00
|
|
|
a ready-to-execute or already-partially-executed statement, with any
|
2017-10-09 03:44:17 +02:00
|
|
|
missing parameter values filled in. (For <command>SELECT</command> statements,
|
2003-05-07 23:46:15 +02:00
|
|
|
a portal is equivalent to an open cursor, but we choose to use a different
|
2017-10-09 03:44:17 +02:00
|
|
|
term since cursors don't handle non-<command>SELECT</command> statements.)
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The overall execution cycle consists of a <firstterm>parse</firstterm> step,
|
2003-04-16 00:51:18 +02:00
|
|
|
which creates a prepared statement from a textual query string; a
|
2017-10-09 03:44:17 +02:00
|
|
|
<firstterm>bind</firstterm> step, which creates a portal given a prepared
|
2003-04-16 00:51:18 +02:00
|
|
|
statement and values for any needed parameters; and an
|
2017-10-09 03:44:17 +02:00
|
|
|
<firstterm>execute</firstterm> step that runs a portal's query. In the case of
|
|
|
|
a query that returns rows (<command>SELECT</command>, <command>SHOW</command>, etc),
|
2003-05-06 23:51:42 +02:00
|
|
|
the execute step can be told to fetch only
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
a limited number of rows, so that multiple execute steps might be needed
|
2003-04-16 00:51:18 +02:00
|
|
|
to complete the operation.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The backend can keep track of multiple prepared statements and portals
|
|
|
|
(but note that these exist only within a session, and are never shared
|
|
|
|
across sessions). Existing prepared statements and portals are
|
|
|
|
referenced by names assigned when they were created. In addition,
|
2017-10-09 03:44:17 +02:00
|
|
|
an <quote>unnamed</quote> prepared statement and portal exist. Although these
|
2003-05-07 01:10:04 +02:00
|
|
|
behave largely the same as named objects, operations on them are optimized
|
|
|
|
for the case of executing a query only once and then discarding it,
|
|
|
|
whereas operations on named objects are optimized on the expectation
|
|
|
|
of multiple uses.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
</sect2>
|
2003-05-07 23:46:15 +02:00
|
|
|
|
|
|
|
<sect2 id="protocol-format-codes">
|
|
|
|
<title>Formats and Format Codes</title>
|
|
|
|
|
|
|
|
<para>
|
2003-11-01 02:56:29 +01:00
|
|
|
Data of a particular data type might be transmitted in any of several
|
2017-10-09 03:44:17 +02:00
|
|
|
different <firstterm>formats</firstterm>. As of <productname>PostgreSQL</productname> 7.4
|
|
|
|
the only supported formats are <quote>text</quote> and <quote>binary</quote>,
|
2003-05-07 23:46:15 +02:00
|
|
|
but the protocol makes provision for future extensions. The desired
|
2017-10-09 03:44:17 +02:00
|
|
|
format for any value is specified by a <firstterm>format code</firstterm>.
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
Clients can specify a format code for each transmitted parameter value
|
2003-05-07 23:46:15 +02:00
|
|
|
and for each column of a query result. Text has format code zero,
|
|
|
|
binary has format code one, and all other format codes are reserved
|
|
|
|
for future definition.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The text representation of values is whatever strings are produced
|
|
|
|
and accepted by the input/output conversion functions for the
|
2003-11-01 02:56:29 +01:00
|
|
|
particular data type. In the transmitted representation, there is
|
2003-05-07 23:46:15 +02:00
|
|
|
no trailing null character; the frontend must add one to received
|
|
|
|
values if it wants to process them as C strings.
|
|
|
|
(The text format does not allow embedded nulls, by the way.)
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Binary representations for integers use network byte order (most
|
2003-11-01 02:56:29 +01:00
|
|
|
significant byte first). For other data types consult the documentation
|
2003-05-07 23:46:15 +02:00
|
|
|
or source code to learn about the binary representation. Keep in mind
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
that binary representations for complex data types might change across
|
2003-05-07 23:46:15 +02:00
|
|
|
server versions; the text format is usually the more portable choice.
|
|
|
|
</para>
|
|
|
|
</sect2>
|
2001-06-23 01:27:48 +02:00
|
|
|
</sect1>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<sect1 id="protocol-flow">
|
|
|
|
<title>Message Flow</title>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
This section describes the message flow and the semantics of each
|
2003-05-07 23:46:15 +02:00
|
|
|
message type. (Details of the exact representation of each message
|
2017-11-23 15:39:47 +01:00
|
|
|
appear in <xref linkend="protocol-message-formats"/>.) There are
|
2003-12-14 01:10:32 +01:00
|
|
|
several different sub-protocols depending on the state of the
|
|
|
|
connection: start-up, query, function call,
|
|
|
|
<command>COPY</command>, and termination. There are also special
|
|
|
|
provisions for asynchronous operations (including notification
|
|
|
|
responses and command cancellation), which can occur at any time
|
|
|
|
after the start-up phase.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<sect2>
|
2011-02-01 23:00:26 +01:00
|
|
|
<title>Start-up</title>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
<para>
|
|
|
|
To begin a session, a frontend opens a connection to the server and sends
|
|
|
|
a startup message. This message includes the names of the user and of the
|
|
|
|
database the user wants to connect to; it also identifies the particular
|
2003-05-07 23:46:15 +02:00
|
|
|
protocol version to be used. (Optionally, the startup message can include
|
|
|
|
additional settings for run-time parameters.)
|
|
|
|
The server then uses this information and
|
2003-04-16 00:51:18 +02:00
|
|
|
the contents of its configuration files (such as
|
|
|
|
<filename>pg_hba.conf</filename>) to determine
|
|
|
|
whether the connection is provisionally acceptable, and what additional
|
|
|
|
authentication is required (if any).
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The server then sends an appropriate authentication request message,
|
|
|
|
to which the frontend must reply with an appropriate authentication
|
|
|
|
response message (such as a password).
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
For all authentication methods except GSSAPI, SSPI and SASL, there is at
|
|
|
|
most one request and one response. In some methods, no response
|
2003-04-16 00:51:18 +02:00
|
|
|
at all is needed from the frontend, and so no authentication request
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
occurs. For GSSAPI, SSPI and SASL, multiple exchanges of packets may be
|
|
|
|
needed to complete the authentication.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The authentication cycle ends with the server either rejecting the
|
2003-09-20 22:12:05 +02:00
|
|
|
connection attempt (ErrorResponse), or sending AuthenticationOk.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The possible messages from the server in this phase are:
|
2001-06-23 01:27:48 +02:00
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>ErrorResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The connection attempt has been rejected.
|
2001-06-23 01:27:48 +02:00
|
|
|
The server then immediately closes the connection.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationOk</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The authentication exchange is successfully completed.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>AuthenticationKerberosV5</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The frontend must now take part in a Kerberos V5
|
2001-06-23 01:27:48 +02:00
|
|
|
authentication dialog (not described here, part of the
|
|
|
|
Kerberos specification) with the server. If this is
|
|
|
|
successful, the server responds with an AuthenticationOk,
|
2014-01-15 17:24:01 +01:00
|
|
|
otherwise it responds with an ErrorResponse. This is no
|
2014-12-17 09:59:21 +01:00
|
|
|
longer supported.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationCleartextPassword</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The frontend must now send a PasswordMessage containing the
|
2005-06-22 17:19:43 +02:00
|
|
|
password in clear-text form. If
|
2001-06-23 01:27:48 +02:00
|
|
|
this is the correct password, the server responds with an
|
|
|
|
AuthenticationOk, otherwise it responds with an ErrorResponse.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationMD5Password</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The frontend must now send a PasswordMessage containing the
|
2015-09-11 03:22:21 +02:00
|
|
|
password (with user name) encrypted via MD5, then encrypted
|
2011-10-14 02:48:50 +02:00
|
|
|
again using the 4-byte random salt specified in the
|
|
|
|
AuthenticationMD5Password message. If this is the correct
|
|
|
|
password, the server responds with an AuthenticationOk,
|
|
|
|
otherwise it responds with an ErrorResponse. The actual
|
|
|
|
PasswordMessage can be computed in SQL as <literal>concat('md5',
|
2017-10-09 03:44:17 +02:00
|
|
|
md5(concat(md5(concat(password, username)), random-salt)))</literal>.
|
|
|
|
(Keep in mind the <function>md5()</function> function returns its
|
2011-10-14 02:48:50 +02:00
|
|
|
result as a hex string.)
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationSCMCredential</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
This response is only possible for local Unix-domain connections
|
2005-06-22 17:19:43 +02:00
|
|
|
on platforms that support SCM credential messages. The frontend
|
|
|
|
must issue an SCM credential message and then send a single data
|
|
|
|
byte. (The contents of the data byte are uninteresting; it's
|
|
|
|
only used to ensure that the server waits long enough to receive
|
|
|
|
the credential message.) If the credential is acceptable,
|
|
|
|
the server responds with an
|
2001-09-21 22:31:49 +02:00
|
|
|
AuthenticationOk, otherwise it responds with an ErrorResponse.
|
Replace use of credential control messages with getsockopt(LOCAL_PEERCRED).
It turns out the reason we hadn't found out about the portability issues
with our credential-control-message code is that almost no modern platforms
use that code at all; the ones that used to need it now offer getpeereid(),
which we choose first. The last holdout was NetBSD, and they added
getpeereid() as of 5.0. So far as I can tell, the only live platform on
which that code was being exercised was Debian/kFreeBSD, ie, FreeBSD kernel
with Linux userland --- since glibc doesn't provide getpeereid(), we fell
back to the control message code. However, the FreeBSD kernel provides a
LOCAL_PEERCRED socket parameter that's functionally equivalent to Linux's
SO_PEERCRED. That is both much simpler to use than control messages, and
superior because it doesn't require receiving a message from the other end
at just the right time.
Therefore, add code to use LOCAL_PEERCRED when necessary, and rip out all
the credential-control-message code in the backend. (libpq still has such
code so that it can still talk to pre-9.1 servers ... but eventually we can
get rid of it there too.) Clean up related autoconf probes, too.
This means that libpq's requirepeer parameter now works on exactly the same
platforms where the backend supports peer authentication, so adjust the
documentation accordingly.
2011-05-31 22:10:46 +02:00
|
|
|
(This message type is only issued by pre-9.1 servers. It may
|
|
|
|
eventually be removed from the protocol specification.)
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2007-07-18 14:00:47 +02:00
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationGSS</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The frontend must now initiate a GSSAPI negotiation. The frontend
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
will send a GSSResponse message with the first part of the GSSAPI
|
2007-07-18 14:00:47 +02:00
|
|
|
data stream in response to this. If further messages are needed,
|
|
|
|
the server will respond with AuthenticationGSSContinue.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2007-12-03 14:40:11 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationSSPI</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2021-06-11 03:38:04 +02:00
|
|
|
The frontend must now initiate an SSPI negotiation. The frontend
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
will send a GSSResponse with the first part of the SSPI
|
2007-12-03 14:40:11 +01:00
|
|
|
data stream in response to this. If further messages are needed,
|
|
|
|
the server will respond with AuthenticationGSSContinue.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
|
|
|
|
</varlistentry>
|
2007-07-18 14:00:47 +02:00
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationGSSContinue</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
This message contains the response data from the previous step
|
2007-12-03 14:40:11 +01:00
|
|
|
of GSSAPI or SSPI negotiation (AuthenticationGSS, AuthenticationSSPI
|
2010-02-16 23:34:57 +01:00
|
|
|
or a previous AuthenticationGSSContinue). If the GSSAPI
|
2007-12-03 14:40:11 +01:00
|
|
|
or SSPI data in this message
|
2007-07-18 14:00:47 +02:00
|
|
|
indicates more data is needed to complete the authentication,
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
the frontend must send that data as another GSSResponse message. If
|
2008-02-08 19:18:05 +01:00
|
|
|
GSSAPI or SSPI authentication is completed by this message, the server
|
|
|
|
will next send AuthenticationOk to indicate successful authentication
|
2007-07-18 14:00:47 +02:00
|
|
|
or ErrorResponse to indicate failure.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationSASL</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
The frontend must now initiate a SASL negotiation, using one of the
|
|
|
|
SASL mechanisms listed in the message. The frontend will send a
|
|
|
|
SASLInitialResponse with the name of the selected mechanism, and the
|
|
|
|
first part of the SASL data stream in response to this. If further
|
|
|
|
messages are needed, the server will respond with
|
2017-11-23 15:39:47 +01:00
|
|
|
AuthenticationSASLContinue. See <xref linkend="sasl-authentication"/>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
for details.
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationSASLContinue</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
This message contains challenge data from the previous step of SASL
|
|
|
|
negotiation (AuthenticationSASL, or a previous
|
|
|
|
AuthenticationSASLContinue). The frontend must respond with a
|
|
|
|
SASLResponse message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>AuthenticationSASLFinal</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
SASL authentication has completed with additional mechanism-specific
|
|
|
|
data for the client. The server will next send AuthenticationOk to
|
|
|
|
indicate successful authentication, or an ErrorResponse to indicate
|
|
|
|
failure. This message is sent only if the SASL mechanism specifies
|
|
|
|
additional data to be sent from server to client at completion.
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2017-11-21 19:56:24 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>NegotiateProtocolVersion</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The server does not support the minor protocol version requested
|
|
|
|
by the client, but does support an earlier version of the protocol;
|
|
|
|
this message indicates the highest supported minor version. This
|
|
|
|
message will also be sent if the client requested unsupported protocol
|
2020-09-01 00:33:37 +02:00
|
|
|
options (i.e., beginning with <literal>_pq_.</literal>) in the
|
2017-11-21 19:56:24 +01:00
|
|
|
startup packet. This message will be followed by an ErrorResponse or
|
|
|
|
a message indicating the success or failure of authentication.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</variablelist>
|
|
|
|
</para>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
|
|
|
<para>
|
|
|
|
If the frontend does not support the authentication method
|
|
|
|
requested by the server, then it should immediately close the
|
|
|
|
connection.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
After having received AuthenticationOk, the frontend must wait
|
|
|
|
for further messages from the server. In this phase a backend process
|
|
|
|
is being started, and the frontend is just an interested bystander.
|
|
|
|
It is still possible for the startup attempt
|
2017-11-21 19:56:24 +01:00
|
|
|
to fail (ErrorResponse) or the server to decline support for the requested
|
|
|
|
minor protocol version (NegotiateProtocolVersion), but in the normal case
|
|
|
|
the backend will send some ParameterStatus messages, BackendKeyData, and
|
|
|
|
finally ReadyForQuery.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
2003-05-07 23:46:15 +02:00
|
|
|
<para>
|
|
|
|
During this phase the backend will attempt to apply any additional
|
|
|
|
run-time parameter settings that were given in the startup message.
|
|
|
|
If successful, these values become session defaults. An error causes
|
|
|
|
ErrorResponse and exit.
|
|
|
|
</para>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
|
|
|
The possible messages from the backend in this phase are:
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>BackendKeyData</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
This message provides secret-key data that the frontend must
|
|
|
|
save if it wants to be able to issue cancel requests later.
|
|
|
|
The frontend should not respond to this message, but should
|
|
|
|
continue listening for a ReadyForQuery message.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>ParameterStatus</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
This message informs the frontend about the current (initial)
|
2004-03-09 17:57:47 +01:00
|
|
|
setting of backend parameters, such as <xref
|
2017-11-23 15:39:47 +01:00
|
|
|
linkend="guc-client-encoding"/> or <xref linkend="guc-datestyle"/>.
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
The frontend can ignore this message, or record the settings
|
2017-11-23 15:39:47 +01:00
|
|
|
for its future use; see <xref linkend="protocol-async"/> for
|
2004-03-09 17:57:47 +01:00
|
|
|
more details. The frontend should not respond to this
|
|
|
|
message, but should continue listening for a ReadyForQuery
|
|
|
|
message.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>ReadyForQuery</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
Start-up is completed. The frontend can now issue commands.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>ErrorResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
Start-up failed. The connection is closed after sending this
|
|
|
|
message.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>NoticeResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
A warning message has been issued. The frontend should
|
|
|
|
display the message but continue listening for ReadyForQuery
|
|
|
|
or ErrorResponse.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
|
|
|
<para>
|
|
|
|
The ReadyForQuery message is the same one that the backend will
|
2003-04-16 00:51:18 +02:00
|
|
|
issue after each command cycle. Depending on the coding needs of
|
2001-06-23 01:27:48 +02:00
|
|
|
the frontend, it is reasonable to consider ReadyForQuery as
|
2003-04-16 00:51:18 +02:00
|
|
|
starting a command cycle, or to consider ReadyForQuery as ending the
|
|
|
|
start-up phase and each subsequent command cycle.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
</sect2>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<sect2>
|
|
|
|
<title>Simple Query</title>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
A simple query cycle is initiated by the frontend sending a Query message
|
2003-05-07 23:46:15 +02:00
|
|
|
to the backend. The message includes an SQL command (or commands)
|
|
|
|
expressed as a text string.
|
|
|
|
The backend then sends one or more response
|
2001-06-23 01:27:48 +02:00
|
|
|
messages depending on the contents of the query command string,
|
|
|
|
and finally a ReadyForQuery response message. ReadyForQuery
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
informs the frontend that it can safely send a new command.
|
2003-04-16 00:51:18 +02:00
|
|
|
(It is not actually necessary for the frontend to wait for
|
|
|
|
ReadyForQuery before issuing another command, but the frontend must
|
|
|
|
then take responsibility for figuring out what happens if the earlier
|
|
|
|
command fails and already-issued later commands succeed.)
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
The possible response messages from the backend are:
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>CommandComplete</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
An SQL command completed normally.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>CopyInResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
The backend is ready to copy data from the frontend to a
|
2017-11-23 15:39:47 +01:00
|
|
|
table; see <xref linkend="protocol-copy"/>.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>CopyOutResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
The backend is ready to copy data from a table to the
|
2017-11-23 15:39:47 +01:00
|
|
|
frontend; see <xref linkend="protocol-copy"/>.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>RowDescription</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-11-22 02:22:10 +01:00
|
|
|
Indicates that rows are about to be returned in response to
|
2005-06-22 17:19:43 +02:00
|
|
|
a <command>SELECT</command>, <command>FETCH</command>, etc query.
|
|
|
|
The contents of this message describe the column layout of the rows.
|
|
|
|
This will be followed by a DataRow message for each row being returned
|
2001-11-22 02:22:10 +01:00
|
|
|
to the frontend.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>DataRow</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
One of the set of rows returned by
|
2005-06-22 17:19:43 +02:00
|
|
|
a <command>SELECT</command>, <command>FETCH</command>, etc query.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>EmptyQueryResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-11-22 02:22:10 +01:00
|
|
|
An empty query string was recognized.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>ErrorResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
An error has occurred.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>ReadyForQuery</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
Processing of the query string is complete. A separate
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
message is sent to indicate this because the query string might
|
2003-04-16 00:51:18 +02:00
|
|
|
contain multiple SQL commands. (CommandComplete marks the
|
2001-06-23 01:27:48 +02:00
|
|
|
end of processing one SQL command, not the whole string.)
|
|
|
|
ReadyForQuery will always be sent, whether processing
|
|
|
|
terminates successfully or with an error.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>NoticeResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
A warning message has been issued in relation to the query.
|
|
|
|
Notices are in addition to other responses, i.e., the backend
|
|
|
|
will continue processing the command.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</variablelist>
|
|
|
|
</para>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The response to a <command>SELECT</command> query (or other queries that
|
|
|
|
return row sets, such as <command>EXPLAIN</command> or <command>SHOW</command>)
|
2003-04-16 00:51:18 +02:00
|
|
|
normally consists of RowDescription, zero or more
|
2003-05-07 23:46:15 +02:00
|
|
|
DataRow messages, and then CommandComplete.
|
2017-10-09 03:44:17 +02:00
|
|
|
<command>COPY</command> to or from the frontend invokes special protocol
|
2017-11-23 15:39:47 +01:00
|
|
|
as described in <xref linkend="protocol-copy"/>.
|
2001-11-22 02:22:10 +01:00
|
|
|
All other query types normally produce only
|
2003-04-16 00:51:18 +02:00
|
|
|
a CommandComplete message.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2001-11-22 02:22:10 +01:00
|
|
|
Since a query string could contain several queries (separated by
|
|
|
|
semicolons), there might be several such response sequences before the
|
|
|
|
backend finishes processing the query string. ReadyForQuery is issued
|
|
|
|
when the entire string has been processed and the backend is ready to
|
|
|
|
accept a new query string.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2001-11-22 02:22:10 +01:00
|
|
|
If a completely empty (no contents other than whitespace) query string
|
|
|
|
is received, the response is EmptyQueryResponse followed by ReadyForQuery.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2001-11-22 02:22:10 +01:00
|
|
|
In the event of an error, ErrorResponse is issued followed by
|
|
|
|
ReadyForQuery. All further processing of the query string is aborted by
|
|
|
|
ErrorResponse (even if more queries remained in it). Note that this
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
might occur partway through the sequence of messages generated by an
|
2001-11-22 02:22:10 +01:00
|
|
|
individual query.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
2003-05-07 23:46:15 +02:00
|
|
|
<para>
|
|
|
|
In simple Query mode, the format of retrieved values is always text,
|
2017-10-09 03:44:17 +02:00
|
|
|
except when the given command is a <command>FETCH</command> from a cursor
|
|
|
|
declared with the <literal>BINARY</literal> option. In that case, the
|
2003-05-07 23:46:15 +02:00
|
|
|
retrieved values are in binary format. The format codes given in
|
|
|
|
the RowDescription message tell which format is being used.
|
|
|
|
</para>
|
|
|
|
|
2001-06-23 01:27:48 +02:00
|
|
|
<para>
|
|
|
|
A frontend must be prepared to accept ErrorResponse and
|
|
|
|
NoticeResponse messages whenever it is expecting any other type of
|
2017-11-23 15:39:47 +01:00
|
|
|
message. See also <xref linkend="protocol-async"/> concerning messages
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
that the backend might generate due to outside events.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
|
|
|
Recommended practice is to code frontends in a state-machine style
|
|
|
|
that will accept any message type at any time that it could make sense,
|
|
|
|
rather than wiring in assumptions about the exact sequence of messages.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
2017-09-07 20:04:41 +02:00
|
|
|
|
|
|
|
<sect3 id="protocol-flow-multi-statement">
|
|
|
|
<title>Multiple Statements in a Simple Query</title>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
When a simple Query message contains more than one SQL statement
|
|
|
|
(separated by semicolons), those statements are executed as a single
|
|
|
|
transaction, unless explicit transaction control commands are included
|
|
|
|
to force a different behavior. For example, if the message contains
|
|
|
|
<programlisting>
|
|
|
|
INSERT INTO mytable VALUES(1);
|
|
|
|
SELECT 1/0;
|
|
|
|
INSERT INTO mytable VALUES(2);
|
|
|
|
</programlisting>
|
2017-10-09 03:44:17 +02:00
|
|
|
then the divide-by-zero failure in the <command>SELECT</command> will force
|
|
|
|
rollback of the first <command>INSERT</command>. Furthermore, because
|
2017-09-07 20:04:41 +02:00
|
|
|
execution of the message is abandoned at the first error, the second
|
2017-10-09 03:44:17 +02:00
|
|
|
<command>INSERT</command> is never attempted at all.
|
2017-09-07 20:04:41 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
If instead the message contains
|
|
|
|
<programlisting>
|
|
|
|
BEGIN;
|
|
|
|
INSERT INTO mytable VALUES(1);
|
|
|
|
COMMIT;
|
|
|
|
INSERT INTO mytable VALUES(2);
|
|
|
|
SELECT 1/0;
|
|
|
|
</programlisting>
|
2017-10-09 03:44:17 +02:00
|
|
|
then the first <command>INSERT</command> is committed by the
|
|
|
|
explicit <command>COMMIT</command> command. The second <command>INSERT</command>
|
|
|
|
and the <command>SELECT</command> are still treated as a single transaction,
|
2017-09-07 20:04:41 +02:00
|
|
|
so that the divide-by-zero failure will roll back the
|
2017-10-09 03:44:17 +02:00
|
|
|
second <command>INSERT</command>, but not the first one.
|
2017-09-07 20:04:41 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
This behavior is implemented by running the statements in a
|
|
|
|
multi-statement Query message in an <firstterm>implicit transaction
|
2017-10-09 03:44:17 +02:00
|
|
|
block</firstterm> unless there is some explicit transaction block for them to
|
2017-09-07 20:04:41 +02:00
|
|
|
run in. The main difference between an implicit transaction block and
|
|
|
|
a regular one is that an implicit block is closed automatically at the
|
|
|
|
end of the Query message, either by an implicit commit if there was no
|
|
|
|
error, or an implicit rollback if there was an error. This is similar
|
|
|
|
to the implicit commit or rollback that happens for a statement
|
|
|
|
executed by itself (when not in a transaction block).
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
If the session is already in a transaction block, as a result of
|
2017-10-09 03:44:17 +02:00
|
|
|
a <command>BEGIN</command> in some previous message, then the Query message
|
2017-09-07 20:04:41 +02:00
|
|
|
simply continues that transaction block, whether the message contains
|
|
|
|
one statement or several. However, if the Query message contains
|
2017-10-09 03:44:17 +02:00
|
|
|
a <command>COMMIT</command> or <command>ROLLBACK</command> closing the existing
|
2017-09-07 20:04:41 +02:00
|
|
|
transaction block, then any following statements are executed in an
|
|
|
|
implicit transaction block.
|
2017-10-09 03:44:17 +02:00
|
|
|
Conversely, if a <command>BEGIN</command> appears in a multi-statement Query
|
2017-09-07 20:04:41 +02:00
|
|
|
message, then it starts a regular transaction block that will only be
|
2017-10-09 03:44:17 +02:00
|
|
|
terminated by an explicit <command>COMMIT</command> or <command>ROLLBACK</command>,
|
2017-09-07 20:04:41 +02:00
|
|
|
whether that appears in this Query message or a later one.
|
2017-10-09 03:44:17 +02:00
|
|
|
If the <command>BEGIN</command> follows some statements that were executed as
|
2017-09-07 20:04:41 +02:00
|
|
|
an implicit transaction block, those statements are not immediately
|
|
|
|
committed; in effect, they are retroactively included into the new
|
|
|
|
regular transaction block.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
A <command>COMMIT</command> or <command>ROLLBACK</command> appearing in an implicit
|
2017-09-07 20:04:41 +02:00
|
|
|
transaction block is executed as normal, closing the implicit block;
|
2017-10-09 03:44:17 +02:00
|
|
|
however, a warning will be issued since a <command>COMMIT</command>
|
|
|
|
or <command>ROLLBACK</command> without a previous <command>BEGIN</command> might
|
2017-09-07 20:04:41 +02:00
|
|
|
represent a mistake. If more statements follow, a new implicit
|
|
|
|
transaction block will be started for them.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Savepoints are not allowed in an implicit transaction block, since
|
|
|
|
they would conflict with the behavior of automatically closing the
|
|
|
|
block upon any error.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Remember that, regardless of any transaction control commands that may
|
|
|
|
be present, execution of the Query message stops at the first error.
|
|
|
|
Thus for example given
|
|
|
|
<programlisting>
|
|
|
|
BEGIN;
|
|
|
|
SELECT 1/0;
|
|
|
|
ROLLBACK;
|
|
|
|
</programlisting>
|
|
|
|
in a single Query message, the session will be left inside a failed
|
2017-10-09 03:44:17 +02:00
|
|
|
regular transaction block, since the <command>ROLLBACK</command> is not
|
|
|
|
reached after the divide-by-zero error. Another <command>ROLLBACK</command>
|
2017-09-07 20:04:41 +02:00
|
|
|
will be needed to restore the session to a usable state.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Another behavior of note is that initial lexical and syntactic
|
|
|
|
analysis is done on the entire query string before any of it is
|
|
|
|
executed. Thus simple errors (such as a misspelled keyword) in later
|
|
|
|
statements can prevent execution of any of the statements. This
|
|
|
|
is normally invisible to users since the statements would all roll
|
|
|
|
back anyway when done as an implicit transaction block. However,
|
|
|
|
it can be visible when attempting to do multiple transactions within a
|
|
|
|
multi-statement Query. For instance, if a typo turned our previous
|
|
|
|
example into
|
|
|
|
<programlisting>
|
|
|
|
BEGIN;
|
|
|
|
INSERT INTO mytable VALUES(1);
|
|
|
|
COMMIT;
|
|
|
|
INSERT INTO mytable VALUES(2);
|
2020-04-11 09:04:57 +02:00
|
|
|
SELCT 1/0;<!-- this typo is intentional -->
|
2017-09-07 20:04:41 +02:00
|
|
|
</programlisting>
|
|
|
|
then none of the statements would get run, resulting in the visible
|
2017-10-09 03:44:17 +02:00
|
|
|
difference that the first <command>INSERT</command> is not committed.
|
2017-09-07 20:04:41 +02:00
|
|
|
Errors detected at semantic analysis or later, such as a misspelled
|
|
|
|
table or column name, do not have this effect.
|
|
|
|
</para>
|
|
|
|
</sect3>
|
2003-04-16 00:51:18 +02:00
|
|
|
</sect2>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2006-01-18 07:49:30 +01:00
|
|
|
<sect2 id="protocol-flow-ext-query">
|
2005-01-23 01:30:59 +01:00
|
|
|
<title>Extended Query</title>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
<para>
|
|
|
|
The extended query protocol breaks down the above-described simple
|
|
|
|
query protocol into multiple steps. The results of preparatory
|
|
|
|
steps can be re-used multiple times for improved efficiency.
|
|
|
|
Furthermore, additional features are available, such as the possibility
|
|
|
|
of supplying data values as separate parameters instead of having to
|
|
|
|
insert them directly into a query string.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
In the extended protocol, the frontend first sends a Parse message,
|
|
|
|
which contains a textual query string, optionally some information
|
2003-11-01 02:56:29 +01:00
|
|
|
about data types of parameter placeholders, and the
|
2003-04-16 00:51:18 +02:00
|
|
|
name of a destination prepared-statement object (an empty string
|
|
|
|
selects the unnamed prepared statement). The response is
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
either ParseComplete or ErrorResponse. Parameter data types can be
|
2003-04-16 00:51:18 +02:00
|
|
|
specified by OID; if not given, the parser attempts to infer the
|
2003-11-01 02:56:29 +01:00
|
|
|
data types in the same way as it would do for untyped literal string
|
2003-04-16 00:51:18 +02:00
|
|
|
constants.
|
|
|
|
</para>
|
|
|
|
|
2005-06-22 17:19:43 +02:00
|
|
|
<note>
|
|
|
|
<para>
|
|
|
|
A parameter data type can be left unspecified by setting it to zero,
|
|
|
|
or by making the array of parameter type OIDs shorter than the
|
2017-10-09 03:44:17 +02:00
|
|
|
number of parameter symbols (<literal>$</literal><replaceable>n</replaceable>)
|
2005-06-22 17:19:43 +02:00
|
|
|
used in the query string. Another special case is that a parameter's
|
2017-10-09 03:44:17 +02:00
|
|
|
type can be specified as <type>void</type> (that is, the OID of the
|
|
|
|
<type>void</type> pseudo-type). This is meant to allow parameter symbols
|
2005-06-22 17:19:43 +02:00
|
|
|
to be used for function parameters that are actually OUT parameters.
|
2017-10-09 03:44:17 +02:00
|
|
|
Ordinarily there is no context in which a <type>void</type> parameter
|
2005-06-22 17:19:43 +02:00
|
|
|
could be used, but if such a parameter symbol appears in a function's
|
|
|
|
parameter list, it is effectively ignored. For example, a function
|
2017-10-09 03:44:17 +02:00
|
|
|
call such as <literal>foo($1,$2,$3,$4)</literal> could match a function with
|
|
|
|
two IN and two OUT arguments, if <literal>$3</literal> and <literal>$4</literal>
|
|
|
|
are specified as having type <type>void</type>.
|
2005-06-22 17:19:43 +02:00
|
|
|
</para>
|
|
|
|
</note>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<note>
|
|
|
|
<para>
|
|
|
|
The query string contained in a Parse message cannot include more
|
|
|
|
than one SQL statement; else a syntax error is reported. This
|
|
|
|
restriction does not exist in the simple-query protocol, but it
|
|
|
|
does exist in the extended protocol, because allowing prepared
|
|
|
|
statements or portals to contain multiple commands would complicate
|
|
|
|
the protocol unduly.
|
|
|
|
</para>
|
|
|
|
</note>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
If successfully created, a named prepared-statement object lasts till
|
|
|
|
the end of the current session, unless explicitly destroyed. An unnamed
|
2003-05-05 02:44:56 +02:00
|
|
|
prepared statement lasts only until the next Parse statement specifying
|
|
|
|
the unnamed statement as destination is issued. (Note that a simple
|
|
|
|
Query message also destroys the unnamed statement.) Named prepared
|
|
|
|
statements must be explicitly closed before they can be redefined by
|
2011-09-16 06:42:53 +02:00
|
|
|
another Parse message, but this is not required for the unnamed statement.
|
2003-04-16 00:51:18 +02:00
|
|
|
Named prepared statements can also be created and accessed at the SQL
|
2017-10-09 03:44:17 +02:00
|
|
|
command level, using <command>PREPARE</command> and <command>EXECUTE</command>.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Once a prepared statement exists, it can be readied for execution using a
|
|
|
|
Bind message. The Bind message gives the name of the source prepared
|
|
|
|
statement (empty string denotes the unnamed prepared statement), the name
|
|
|
|
of the destination portal (empty string denotes the unnamed portal), and
|
|
|
|
the values to use for any parameter placeholders present in the prepared
|
2003-05-07 23:46:15 +02:00
|
|
|
statement. The
|
2003-04-16 00:51:18 +02:00
|
|
|
supplied parameter set must match those needed by the prepared statement.
|
2017-10-09 03:44:17 +02:00
|
|
|
(If you declared any <type>void</type> parameters in the Parse message,
|
2005-06-22 17:19:43 +02:00
|
|
|
pass NULL values for them in the Bind message.)
|
2003-05-07 23:46:15 +02:00
|
|
|
Bind also specifies the format to use for any data returned
|
|
|
|
by the query; the format can be specified overall, or per-column.
|
|
|
|
The response is either BindComplete or ErrorResponse.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
2003-05-07 23:46:15 +02:00
|
|
|
<note>
|
|
|
|
<para>
|
|
|
|
The choice between text and binary output is determined by the format
|
|
|
|
codes given in Bind, regardless of the SQL command involved. The
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>BINARY</literal> attribute in cursor declarations is irrelevant when
|
2003-05-07 23:46:15 +02:00
|
|
|
using extended query protocol.
|
|
|
|
</para>
|
|
|
|
</note>
|
|
|
|
|
2004-06-11 03:09:22 +02:00
|
|
|
<para>
|
2011-09-16 06:42:53 +02:00
|
|
|
Query planning typically occurs when the Bind message is processed.
|
|
|
|
If the prepared statement has no parameters, or is executed repeatedly,
|
|
|
|
the server might save the created plan and re-use it during subsequent
|
|
|
|
Bind messages for the same prepared statement. However, it will do so
|
|
|
|
only if it finds that a generic plan can be created that is not much
|
|
|
|
less efficient than a plan that depends on the specific parameter values
|
|
|
|
supplied. This happens transparently so far as the protocol is concerned.
|
2004-06-11 03:09:22 +02:00
|
|
|
</para>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
2003-05-05 02:44:56 +02:00
|
|
|
If successfully created, a named portal object lasts till the end of the
|
|
|
|
current transaction, unless explicitly destroyed. An unnamed portal is
|
|
|
|
destroyed at the end of the transaction, or as soon as the next Bind
|
|
|
|
statement specifying the unnamed portal as destination is issued. (Note
|
|
|
|
that a simple Query message also destroys the unnamed portal.) Named
|
2011-09-16 06:42:53 +02:00
|
|
|
portals must be explicitly closed before they can be redefined by another
|
|
|
|
Bind message, but this is not required for the unnamed portal.
|
2003-04-16 00:51:18 +02:00
|
|
|
Named portals can also be created and accessed at the SQL
|
2017-10-09 03:44:17 +02:00
|
|
|
command level, using <command>DECLARE CURSOR</command> and <command>FETCH</command>.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Once a portal exists, it can be executed using an Execute message.
|
|
|
|
The Execute message specifies the portal name (empty string denotes the
|
2003-05-07 23:46:15 +02:00
|
|
|
unnamed portal) and
|
2017-10-09 03:44:17 +02:00
|
|
|
a maximum result-row count (zero meaning <quote>fetch all rows</quote>).
|
2003-05-07 23:46:15 +02:00
|
|
|
The result-row count is only meaningful for portals
|
2003-11-01 02:56:29 +01:00
|
|
|
containing commands that return row sets; in other cases the command is
|
2003-05-07 23:46:15 +02:00
|
|
|
always executed to completion, and the row count is ignored.
|
|
|
|
The possible
|
2003-04-16 00:51:18 +02:00
|
|
|
responses to Execute are the same as those described above for queries
|
|
|
|
issued via simple query protocol, except that Execute doesn't cause
|
2004-09-24 01:35:07 +02:00
|
|
|
ReadyForQuery or RowDescription to be issued.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
If Execute terminates before completing the execution of a portal
|
|
|
|
(due to reaching a nonzero result-row count), it will send a
|
|
|
|
PortalSuspended message; the appearance of this message tells the frontend
|
|
|
|
that another Execute should be issued against the same portal to
|
|
|
|
complete the operation. The CommandComplete message indicating
|
2003-04-16 22:53:38 +02:00
|
|
|
completion of the source SQL command is not sent until
|
|
|
|
the portal's execution is completed. Therefore, an Execute phase is
|
|
|
|
always terminated by the appearance of exactly one of these messages:
|
|
|
|
CommandComplete, EmptyQueryResponse (if the portal was created from
|
|
|
|
an empty query string), ErrorResponse, or PortalSuspended.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
At completion of each series of extended-query messages, the frontend
|
|
|
|
should issue a Sync message. This parameterless message causes the
|
|
|
|
backend to close the current transaction if it's not inside a
|
2017-10-09 03:44:17 +02:00
|
|
|
<command>BEGIN</command>/<command>COMMIT</command> transaction block (<quote>close</quote>
|
2003-04-16 00:51:18 +02:00
|
|
|
meaning to commit if no error, or roll back if error). Then a
|
|
|
|
ReadyForQuery response is issued. The purpose of Sync is to provide
|
2003-09-20 22:12:05 +02:00
|
|
|
a resynchronization point for error recovery. When an error is detected
|
2003-04-16 00:51:18 +02:00
|
|
|
while processing any extended-query message, the backend issues
|
|
|
|
ErrorResponse, then reads and discards messages until a Sync is reached,
|
|
|
|
then issues ReadyForQuery and returns to normal message processing.
|
|
|
|
(But note that no skipping occurs if an error is detected
|
2017-10-09 03:44:17 +02:00
|
|
|
<emphasis>while</emphasis> processing Sync — this ensures that there is one
|
2003-04-16 00:51:18 +02:00
|
|
|
and only one ReadyForQuery sent for each Sync.)
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<note>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
Sync does not cause a transaction block opened with <command>BEGIN</command>
|
2003-04-16 00:51:18 +02:00
|
|
|
to be closed. It is possible to detect this situation since the
|
|
|
|
ReadyForQuery message includes transaction status information.
|
|
|
|
</para>
|
|
|
|
</note>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
In addition to these fundamental, required operations, there are several
|
|
|
|
optional operations that can be used with extended-query protocol.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The Describe message (portal variant) specifies the name of an existing
|
|
|
|
portal (or an empty string for the unnamed portal). The response is a
|
|
|
|
RowDescription message describing the rows that will be returned by
|
|
|
|
executing the portal; or a NoData message if the portal does not contain a
|
2003-05-06 23:51:42 +02:00
|
|
|
query that will return rows; or ErrorResponse if there is no such portal.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The Describe message (statement variant) specifies the name of an existing
|
|
|
|
prepared statement (or an empty string for the unnamed prepared
|
|
|
|
statement). The response is a ParameterDescription message describing the
|
2003-05-07 23:46:15 +02:00
|
|
|
parameters needed by the statement, followed by a RowDescription message
|
|
|
|
describing the rows that will be returned when the statement is eventually
|
|
|
|
executed (or a NoData message if the statement will not return rows).
|
|
|
|
ErrorResponse is issued if there is no such prepared statement. Note that
|
|
|
|
since Bind has not yet been issued, the formats to be used for returned
|
|
|
|
columns are not yet known to the backend; the format code fields in the
|
|
|
|
RowDescription message will be zeroes in this case.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
2003-05-06 23:51:42 +02:00
|
|
|
<tip>
|
|
|
|
<para>
|
|
|
|
In most scenarios the frontend should issue one or the other variant
|
|
|
|
of Describe before issuing Execute, to ensure that it knows how to
|
|
|
|
interpret the results it will get back.
|
|
|
|
</para>
|
|
|
|
</tip>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
|
|
|
The Close message closes an existing prepared statement or portal
|
2003-05-05 02:44:56 +02:00
|
|
|
and releases resources. It is not an error to issue Close against
|
|
|
|
a nonexistent statement or portal name. The response is normally
|
|
|
|
CloseComplete, but could be ErrorResponse if some difficulty is
|
|
|
|
encountered while releasing resources. Note that closing a prepared
|
|
|
|
statement implicitly closes any open portals that were constructed
|
|
|
|
from that statement.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The Flush message does not cause any specific output to be generated,
|
|
|
|
but forces the backend to deliver any data pending in its output
|
|
|
|
buffers. A Flush must be sent after any extended-query command except
|
|
|
|
Sync, if the frontend wishes to examine the results of that command before
|
2003-05-05 02:44:56 +02:00
|
|
|
issuing more commands. Without Flush, messages returned by the backend
|
|
|
|
will be combined into the minimum possible number of packets to minimize
|
|
|
|
network overhead.
|
2001-11-22 02:22:10 +01:00
|
|
|
</para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
<note>
|
|
|
|
<para>
|
|
|
|
The simple Query message is approximately equivalent to the series Parse,
|
2003-05-05 02:44:56 +02:00
|
|
|
Bind, portal Describe, Execute, Close, Sync, using the unnamed prepared
|
|
|
|
statement and portal objects and no parameters. One difference is that
|
|
|
|
it will accept multiple SQL statements in the query string, automatically
|
2003-04-16 00:51:18 +02:00
|
|
|
performing the bind/describe/execute sequence for each one in succession.
|
2003-05-05 02:44:56 +02:00
|
|
|
Another difference is that it will not return ParseComplete, BindComplete,
|
|
|
|
CloseComplete, or NoData messages.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
</note>
|
2001-06-23 01:27:48 +02:00
|
|
|
</sect2>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<sect2>
|
|
|
|
<title>Function Call</title>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2003-05-07 23:46:15 +02:00
|
|
|
<para>
|
|
|
|
The Function Call sub-protocol allows the client to request a direct
|
|
|
|
call of any function that exists in the database's
|
|
|
|
<structname>pg_proc</structname> system catalog. The client must have
|
|
|
|
execute permission for the function.
|
|
|
|
</para>
|
|
|
|
|
2003-04-16 22:53:38 +02:00
|
|
|
<note>
|
|
|
|
<para>
|
|
|
|
The Function Call sub-protocol is a legacy feature that is probably best
|
|
|
|
avoided in new code. Similar results can be accomplished by setting up
|
2017-10-09 03:44:17 +02:00
|
|
|
a prepared statement that does <literal>SELECT function($1, ...)</literal>.
|
2003-04-16 22:53:38 +02:00
|
|
|
The Function Call cycle can then be replaced with Bind/Execute.
|
|
|
|
</para>
|
|
|
|
</note>
|
|
|
|
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
A Function Call cycle is initiated by the frontend sending a
|
|
|
|
FunctionCall message to the backend. The backend then sends one
|
|
|
|
or more response messages depending on the results of the function
|
|
|
|
call, and finally a ReadyForQuery response message. ReadyForQuery
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
informs the frontend that it can safely send a new query or
|
2001-06-23 01:27:48 +02:00
|
|
|
function call.
|
|
|
|
</para>
|
|
|
|
|
2003-04-16 22:53:38 +02:00
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
The possible response messages from the backend are:
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>ErrorResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
An error has occurred.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>FunctionCallResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The function call was completed and returned the result given
|
2005-06-22 17:19:43 +02:00
|
|
|
in the message.
|
|
|
|
(Note that the Function Call protocol can only handle a single
|
|
|
|
scalar result, not a row type or set of results.)
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>ReadyForQuery</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
Processing of the function call is complete. ReadyForQuery
|
|
|
|
will always be sent, whether processing terminates
|
|
|
|
successfully or with an error.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>NoticeResponse</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
A warning message has been issued in relation to the function
|
|
|
|
call. Notices are in addition to other responses, i.e., the
|
|
|
|
backend will continue processing the command.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
2003-04-16 00:51:18 +02:00
|
|
|
</sect2>
|
|
|
|
|
|
|
|
<sect2 id="protocol-copy">
|
|
|
|
<title>COPY Operations</title>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The <command>COPY</command> command allows high-speed bulk data transfer
|
2003-04-16 00:51:18 +02:00
|
|
|
to or from the server. Copy-in and copy-out operations each switch
|
|
|
|
the connection into a distinct sub-protocol, which lasts until the
|
|
|
|
operation is completed.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Copy-in mode (data transfer to the server) is initiated when the
|
2017-10-09 03:44:17 +02:00
|
|
|
backend executes a <command>COPY FROM STDIN</command> SQL statement. The backend
|
2003-04-16 00:51:18 +02:00
|
|
|
sends a CopyInResponse message to the frontend. The frontend should
|
2003-04-19 02:02:30 +02:00
|
|
|
then send zero or more CopyData messages, forming a stream of input
|
|
|
|
data. (The message boundaries are not required to have anything to do
|
|
|
|
with row boundaries, although that is often a reasonable choice.)
|
2003-04-16 00:51:18 +02:00
|
|
|
The frontend can terminate the copy-in mode by sending either a CopyDone
|
|
|
|
message (allowing successful termination) or a CopyFail message (which
|
2017-10-09 03:44:17 +02:00
|
|
|
will cause the <command>COPY</command> SQL statement to fail with an
|
2003-04-16 00:51:18 +02:00
|
|
|
error). The backend then reverts to the command-processing mode it was
|
2017-10-09 03:44:17 +02:00
|
|
|
in before the <command>COPY</command> started, which will be either simple or
|
2003-04-16 22:53:38 +02:00
|
|
|
extended query protocol. It will next send either CommandComplete
|
|
|
|
(if successful) or ErrorResponse (if not).
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
In the event of a backend-detected error during copy-in mode (including
|
2010-02-16 23:34:57 +01:00
|
|
|
receipt of a CopyFail message), the backend will issue an ErrorResponse
|
2017-10-09 03:44:17 +02:00
|
|
|
message. If the <command>COPY</command> command was issued via an extended-query
|
2003-04-16 00:51:18 +02:00
|
|
|
message, the backend will now discard frontend messages until a Sync
|
|
|
|
message is received, then it will issue ReadyForQuery and return to normal
|
2017-10-09 03:44:17 +02:00
|
|
|
processing. If the <command>COPY</command> command was issued in a simple
|
2003-04-16 00:51:18 +02:00
|
|
|
Query message, the rest of that message is discarded and ReadyForQuery
|
2003-04-19 02:02:30 +02:00
|
|
|
is issued. In either case, any subsequent CopyData, CopyDone, or CopyFail
|
|
|
|
messages issued by the frontend will simply be dropped.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
2003-08-13 20:56:21 +02:00
|
|
|
<para>
|
|
|
|
The backend will ignore Flush and Sync messages received during copy-in
|
|
|
|
mode. Receipt of any other non-copy message type constitutes an error
|
|
|
|
that will abort the copy-in state as described above. (The exception for
|
|
|
|
Flush and Sync is for the convenience of client libraries that always
|
|
|
|
send Flush or Sync after an Execute message, without checking whether
|
2017-10-09 03:44:17 +02:00
|
|
|
the command to be executed is a <command>COPY FROM STDIN</command>.)
|
2003-08-13 20:56:21 +02:00
|
|
|
</para>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
|
|
|
Copy-out mode (data transfer from the server) is initiated when the
|
2017-10-09 03:44:17 +02:00
|
|
|
backend executes a <command>COPY TO STDOUT</command> SQL statement. The backend
|
2003-04-16 00:51:18 +02:00
|
|
|
sends a CopyOutResponse message to the frontend, followed by
|
2003-04-19 02:02:30 +02:00
|
|
|
zero or more CopyData messages (always one per row), followed by CopyDone.
|
2003-04-16 00:51:18 +02:00
|
|
|
The backend then reverts to the command-processing mode it was
|
2017-10-09 03:44:17 +02:00
|
|
|
in before the <command>COPY</command> started, and sends CommandComplete.
|
2003-05-07 01:10:04 +02:00
|
|
|
The frontend cannot abort the transfer (except by closing the connection
|
|
|
|
or issuing a Cancel request),
|
2003-04-19 02:02:30 +02:00
|
|
|
but it can discard unwanted CopyData and CopyDone messages.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
In the event of a backend-detected error during copy-out mode,
|
|
|
|
the backend will issue an ErrorResponse message and revert to normal
|
2008-01-14 19:46:17 +01:00
|
|
|
processing. The frontend should treat receipt of ErrorResponse as
|
|
|
|
terminating the copy-out mode.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2008-01-15 23:18:20 +01:00
|
|
|
It is possible for NoticeResponse and ParameterStatus messages to be
|
|
|
|
interspersed between CopyData messages; frontends must handle these cases,
|
|
|
|
and should be prepared for other asynchronous message types as well (see
|
2017-11-23 15:39:47 +01:00
|
|
|
<xref linkend="protocol-async"/>). Otherwise, any message type other than
|
2008-01-14 19:46:17 +01:00
|
|
|
CopyData or CopyDone may be treated as terminating copy-out mode.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
2003-05-07 23:46:15 +02:00
|
|
|
|
|
|
|
<para>
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
There is another Copy-related mode called copy-both, which allows
|
2017-10-09 03:44:17 +02:00
|
|
|
high-speed bulk data transfer to <emphasis>and</emphasis> from the server.
|
2010-12-11 15:27:37 +01:00
|
|
|
Copy-both mode is initiated when a backend in walsender mode
|
|
|
|
executes a <command>START_REPLICATION</command> statement. The
|
|
|
|
backend sends a CopyBothResponse message to the frontend. Both
|
|
|
|
the backend and the frontend may then send CopyData messages
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
until either end sends a CopyDone message. After the client
|
|
|
|
sends a CopyDone message, the connection goes from copy-both mode to
|
|
|
|
copy-out mode, and the client may not send any more CopyData messages.
|
|
|
|
Similarly, when the server sends a CopyDone message, the connection
|
|
|
|
goes into copy-in mode, and the server may not send any more CopyData
|
|
|
|
messages. After both sides have sent a CopyDone message, the copy mode
|
|
|
|
is terminated, and the backend reverts to the command-processing mode.
|
2013-04-29 12:29:32 +02:00
|
|
|
In the event of a backend-detected error during copy-both mode,
|
|
|
|
the backend will issue an ErrorResponse message, discard frontend messages
|
|
|
|
until a Sync message is received, and then issue ReadyForQuery and return
|
|
|
|
to normal processing. The frontend should treat receipt of ErrorResponse
|
|
|
|
as terminating the copy in both directions; no CopyDone should be sent
|
2017-11-23 15:39:47 +01:00
|
|
|
in this case. See <xref linkend="protocol-replication"/> for more
|
2013-04-29 12:29:32 +02:00
|
|
|
information on the subprotocol transmitted over copy-both mode.
|
2003-05-07 23:46:15 +02:00
|
|
|
</para>
|
2010-12-11 15:27:37 +01:00
|
|
|
|
|
|
|
<para>
|
|
|
|
The CopyInResponse, CopyOutResponse and CopyBothResponse messages
|
|
|
|
include fields that inform the frontend of the number of columns
|
|
|
|
per row and the format codes being used for each column. (As of
|
2017-10-09 03:44:17 +02:00
|
|
|
the present implementation, all columns in a given <command>COPY</command>
|
2010-12-11 15:27:37 +01:00
|
|
|
operation will use the same format, but the message design does not
|
|
|
|
assume this.)
|
|
|
|
</para>
|
|
|
|
|
2001-06-23 01:27:48 +02:00
|
|
|
</sect2>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<sect2 id="protocol-async">
|
|
|
|
<title>Asynchronous Operations</title>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
|
|
|
There are several cases in which the backend will send messages that
|
|
|
|
are not specifically prompted by the frontend's command stream.
|
|
|
|
Frontends must be prepared to deal with these messages at any time,
|
|
|
|
even when not engaged in a query.
|
|
|
|
At minimum, one should check for these cases before beginning to
|
|
|
|
read a query response.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
It is possible for NoticeResponse messages to be generated due to
|
|
|
|
outside activity; for example, if the database administrator commands
|
2017-10-09 03:44:17 +02:00
|
|
|
a <quote>fast</quote> database shutdown, the backend will send a NoticeResponse
|
2003-04-16 00:51:18 +02:00
|
|
|
indicating this fact before closing the connection. Accordingly,
|
|
|
|
frontends should always be prepared to accept and display NoticeResponse
|
|
|
|
messages, even when the connection is nominally idle.
|
|
|
|
</para>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
ParameterStatus messages will be generated whenever the active
|
|
|
|
value changes for any of the parameters the backend believes the
|
|
|
|
frontend should know about. Most commonly this occurs in response
|
2017-10-09 03:44:17 +02:00
|
|
|
to a <command>SET</command> SQL command executed by the frontend, and
|
2004-11-15 07:32:15 +01:00
|
|
|
this case is effectively synchronous — but it is also possible
|
2003-12-14 01:10:32 +01:00
|
|
|
for parameter status changes to occur because the administrator
|
|
|
|
changed a configuration file and then sent the
|
2006-06-18 17:38:37 +02:00
|
|
|
<systemitem>SIGHUP</systemitem> signal to the server. Also,
|
2003-12-14 01:10:32 +01:00
|
|
|
if a <command>SET</command> command is rolled back, an appropriate
|
|
|
|
ParameterStatus message will be generated to report the current
|
|
|
|
effective value.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
At present there is a hard-wired set of parameters for which
|
|
|
|
ParameterStatus will be generated: they are
|
2017-10-09 03:44:17 +02:00
|
|
|
<varname>server_version</varname>,
|
|
|
|
<varname>server_encoding</varname>,
|
|
|
|
<varname>client_encoding</varname>,
|
|
|
|
<varname>application_name</varname>,
|
2021-03-02 19:53:46 +01:00
|
|
|
<varname>default_transaction_read_only</varname>,
|
2021-01-05 22:18:01 +01:00
|
|
|
<varname>in_hot_standby</varname>,
|
2017-10-09 03:44:17 +02:00
|
|
|
<varname>is_superuser</varname>,
|
|
|
|
<varname>session_authorization</varname>,
|
|
|
|
<varname>DateStyle</varname>,
|
|
|
|
<varname>IntervalStyle</varname>,
|
|
|
|
<varname>TimeZone</varname>,
|
|
|
|
<varname>integer_datetimes</varname>, and
|
|
|
|
<varname>standard_conforming_strings</varname>.
|
|
|
|
(<varname>server_encoding</varname>, <varname>TimeZone</varname>, and
|
|
|
|
<varname>integer_datetimes</varname> were not reported by releases before 8.0;
|
|
|
|
<varname>standard_conforming_strings</varname> was not reported by releases
|
2009-12-02 05:54:10 +01:00
|
|
|
before 8.1;
|
2017-10-09 03:44:17 +02:00
|
|
|
<varname>IntervalStyle</varname> was not reported by releases before 8.4;
|
2021-01-05 22:18:01 +01:00
|
|
|
<varname>application_name</varname> was not reported by releases before
|
|
|
|
9.0;
|
2021-03-02 19:53:46 +01:00
|
|
|
<varname>default_transaction_read_only</varname> and
|
|
|
|
<varname>in_hot_standby</varname> were not reported by releases before
|
2021-01-05 22:18:01 +01:00
|
|
|
14.)
|
2004-08-16 04:12:29 +02:00
|
|
|
Note that
|
2017-10-09 03:44:17 +02:00
|
|
|
<varname>server_version</varname>,
|
|
|
|
<varname>server_encoding</varname> and
|
|
|
|
<varname>integer_datetimes</varname>
|
2004-08-16 04:12:29 +02:00
|
|
|
are pseudo-parameters that cannot change after startup.
|
2003-04-16 00:51:18 +02:00
|
|
|
This set might change in the future, or even become configurable.
|
|
|
|
Accordingly, a frontend should simply ignore ParameterStatus for
|
|
|
|
parameters that it does not understand or care about.
|
|
|
|
</para>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
If a frontend issues a <command>LISTEN</command> command, then the
|
|
|
|
backend will send a NotificationResponse message (not to be
|
|
|
|
confused with NoticeResponse!) whenever a
|
|
|
|
<command>NOTIFY</command> command is executed for the same
|
2010-02-16 23:34:57 +01:00
|
|
|
channel name.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
<note>
|
|
|
|
<para>
|
|
|
|
At present, NotificationResponse can only be sent outside a
|
|
|
|
transaction, and thus it will not occur in the middle of a
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
command-response series, though it might occur just before ReadyForQuery.
|
2003-04-16 00:51:18 +02:00
|
|
|
It is unwise to design frontend logic that assumes that, however.
|
|
|
|
Good practice is to be able to accept NotificationResponse at any
|
|
|
|
point in the protocol.
|
|
|
|
</para>
|
|
|
|
</note>
|
2001-06-23 01:27:48 +02:00
|
|
|
</sect2>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<sect2>
|
2011-06-29 08:26:14 +02:00
|
|
|
<title>Canceling Requests in Progress</title>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
During the processing of a query, the frontend might request
|
2001-06-23 01:27:48 +02:00
|
|
|
cancellation of the query. The cancel request is not sent
|
|
|
|
directly on the open connection to the backend for reasons of
|
|
|
|
implementation efficiency: we don't want to have the backend
|
|
|
|
constantly checking for new input from the frontend during query
|
|
|
|
processing. Cancel requests should be relatively infrequent, so
|
|
|
|
we make them slightly cumbersome in order to avoid a penalty in
|
|
|
|
the normal case.
|
|
|
|
</para>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
To issue a cancel request, the frontend opens a new connection to
|
|
|
|
the server and sends a CancelRequest message, rather than the
|
2003-04-16 00:51:18 +02:00
|
|
|
StartupMessage message that would ordinarily be sent across a new
|
2001-06-23 01:27:48 +02:00
|
|
|
connection. The server will process this request and then close
|
|
|
|
the connection. For security reasons, no direct reply is made to
|
|
|
|
the cancel request message.
|
|
|
|
</para>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
A CancelRequest message will be ignored unless it contains the
|
|
|
|
same key data (PID and secret key) passed to the frontend during
|
|
|
|
connection start-up. If the request matches the PID and secret
|
|
|
|
key for a currently executing backend, the processing of the
|
2001-11-22 02:22:10 +01:00
|
|
|
current query is aborted. (In the existing implementation, this is
|
2001-06-23 01:27:48 +02:00
|
|
|
done by sending a special signal to the backend process that is
|
|
|
|
processing the query.)
|
|
|
|
</para>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
The cancellation signal might or might not have any effect — for
|
2001-06-23 01:27:48 +02:00
|
|
|
example, if it arrives after the backend has finished processing
|
|
|
|
the query, then it will have no effect. If the cancellation is
|
|
|
|
effective, it results in the current command being terminated
|
|
|
|
early with an error message.
|
|
|
|
</para>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
The upshot of all this is that for reasons of both security and
|
|
|
|
efficiency, the frontend has no direct way to tell whether a
|
|
|
|
cancel request has succeeded. It must continue to wait for the
|
|
|
|
backend to respond to the query. Issuing a cancel simply improves
|
|
|
|
the odds that the current query will finish soon, and improves the
|
|
|
|
odds that it will fail with an error message instead of
|
|
|
|
succeeding.
|
|
|
|
</para>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
Since the cancel request is sent across a new connection to the
|
|
|
|
server and not across the regular frontend/backend communication
|
|
|
|
link, it is possible for the cancel request to be issued by any
|
|
|
|
process, not just the frontend whose query is to be canceled.
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
This might provide additional flexibility when building
|
2001-06-23 01:27:48 +02:00
|
|
|
multiple-process applications. It also introduces a security
|
|
|
|
risk, in that unauthorized persons might try to cancel queries.
|
|
|
|
The security risk is addressed by requiring a dynamically
|
|
|
|
generated secret key to be supplied in cancel requests.
|
|
|
|
</para>
|
|
|
|
</sect2>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<sect2>
|
|
|
|
<title>Termination</title>
|
2001-06-23 01:27:48 +02:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
The normal, graceful termination procedure is that the frontend
|
|
|
|
sends a Terminate message and immediately closes the connection.
|
2003-04-16 00:51:18 +02:00
|
|
|
On receipt of this message, the backend closes the connection and
|
|
|
|
terminates.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
|
|
|
In rare cases (such as an administrator-commanded database shutdown)
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
the backend might disconnect without any frontend request to do so.
|
2003-04-16 00:51:18 +02:00
|
|
|
In such cases the backend will attempt to send an error or notice message
|
|
|
|
giving the reason for the disconnection before it closes the connection.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Other termination scenarios arise from various failure cases, such as core
|
|
|
|
dump at one end or the other, loss of the communications link, loss of
|
|
|
|
message-boundary synchronization, etc. If either frontend or backend sees
|
|
|
|
an unexpected closure of the connection, it should clean
|
|
|
|
up and terminate. The frontend has the option of launching a new backend
|
|
|
|
by recontacting the server if it doesn't want to terminate itself.
|
|
|
|
Closing the connection is also advisable if an unrecognizable message type
|
|
|
|
is received, since this probably indicates loss of message-boundary sync.
|
2001-06-23 01:27:48 +02:00
|
|
|
</para>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
|
|
|
<para>
|
|
|
|
For either normal or abnormal termination, any open transaction is
|
|
|
|
rolled back, not committed. One should note however that if a
|
2003-12-14 01:10:32 +01:00
|
|
|
frontend disconnects while a non-<command>SELECT</command> query
|
|
|
|
is being processed, the backend will probably finish the query
|
|
|
|
before noticing the disconnection. If the query is outside any
|
2017-10-09 03:44:17 +02:00
|
|
|
transaction block (<command>BEGIN</command> ... <command>COMMIT</command>
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
sequence) then its results might be committed before the
|
2003-12-14 01:10:32 +01:00
|
|
|
disconnection is recognized.
|
2001-11-22 02:22:10 +01:00
|
|
|
</para>
|
|
|
|
</sect2>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<sect2>
|
|
|
|
<title><acronym>SSL</acronym> Session Encryption</title>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
If <productname>PostgreSQL</productname> was built with
|
2003-12-14 01:10:32 +01:00
|
|
|
<acronym>SSL</acronym> support, frontend/backend communications
|
|
|
|
can be encrypted using <acronym>SSL</acronym>. This provides
|
|
|
|
communication security in environments where attackers might be
|
|
|
|
able to capture the session traffic. For more information on
|
|
|
|
encrypting <productname>PostgreSQL</productname> sessions with
|
2017-11-23 15:39:47 +01:00
|
|
|
<acronym>SSL</acronym>, see <xref linkend="ssl-tcp"/>.
|
2001-11-22 02:22:10 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
To initiate an <acronym>SSL</acronym>-encrypted connection, the
|
|
|
|
frontend initially sends an SSLRequest message rather than a
|
|
|
|
StartupMessage. The server then responds with a single byte
|
2017-10-09 03:44:17 +02:00
|
|
|
containing <literal>S</literal> or <literal>N</literal>, indicating that it is
|
2003-12-14 01:10:32 +01:00
|
|
|
willing or unwilling to perform <acronym>SSL</acronym>,
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
respectively. The frontend might close the connection at this point
|
2003-12-14 01:10:32 +01:00
|
|
|
if it is dissatisfied with the response. To continue after
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>S</literal>, perform an <acronym>SSL</acronym> startup handshake
|
2003-12-14 01:10:32 +01:00
|
|
|
(not described here, part of the <acronym>SSL</acronym>
|
|
|
|
specification) with the server. If this is successful, continue
|
|
|
|
with sending the usual StartupMessage. In this case the
|
|
|
|
StartupMessage and all subsequent data will be
|
|
|
|
<acronym>SSL</acronym>-encrypted. To continue after
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>N</literal>, send the usual StartupMessage and proceed without
|
2001-11-22 02:22:10 +01:00
|
|
|
encryption.
|
2020-12-28 23:44:17 +01:00
|
|
|
(Alternatively, it is permissible to issue a GSSENCRequest message
|
|
|
|
after an <literal>N</literal> response to try to
|
|
|
|
use <acronym>GSSAPI</acronym> encryption instead
|
|
|
|
of <acronym>SSL</acronym>.)
|
2001-11-22 02:22:10 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
The frontend should also be prepared to handle an ErrorMessage
|
|
|
|
response to SSLRequest from the server. This would only occur if
|
|
|
|
the server predates the addition of <acronym>SSL</acronym> support
|
2017-10-09 03:44:17 +02:00
|
|
|
to <productname>PostgreSQL</productname>. (Such servers are now very ancient,
|
2011-09-16 06:42:53 +02:00
|
|
|
and likely do not exist in the wild anymore.)
|
|
|
|
In this case the connection must
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
be closed, but the frontend might choose to open a fresh connection
|
2003-12-14 01:10:32 +01:00
|
|
|
and proceed without requesting <acronym>SSL</acronym>.
|
2001-11-22 02:22:10 +01:00
|
|
|
</para>
|
|
|
|
|
2021-11-08 17:14:56 +01:00
|
|
|
<para>
|
|
|
|
When <acronym>SSL</acronym> encryption can be performed, the server
|
|
|
|
is expected to send only the single <literal>S</literal> byte and then
|
|
|
|
wait for the frontend to initiate an <acronym>SSL</acronym> handshake.
|
|
|
|
If additional bytes are available to read at this point, it likely
|
|
|
|
means that a man-in-the-middle is attempting to perform a
|
|
|
|
buffer-stuffing attack
|
|
|
|
(<ulink url="https://www.postgresql.org/support/security/CVE-2021-23222/">CVE-2021-23222</ulink>).
|
|
|
|
Frontends should be coded either to read exactly one byte from the
|
|
|
|
socket before turning the socket over to their SSL library, or to
|
|
|
|
treat it as a protocol violation if they find they have read additional
|
|
|
|
bytes.
|
|
|
|
</para>
|
|
|
|
|
2001-11-22 02:22:10 +01:00
|
|
|
<para>
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
An initial SSLRequest can also be used in a connection that is being
|
2001-11-22 02:22:10 +01:00
|
|
|
opened to send a CancelRequest message.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
While the protocol itself does not provide a way for the server to
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
force <acronym>SSL</acronym> encryption, the administrator can
|
2003-12-14 01:10:32 +01:00
|
|
|
configure the server to reject unencrypted sessions as a byproduct
|
|
|
|
of authentication checking.
|
2001-11-22 02:22:10 +01:00
|
|
|
</para>
|
2001-06-23 01:27:48 +02:00
|
|
|
</sect2>
|
2019-04-20 03:22:22 +02:00
|
|
|
|
|
|
|
<sect2>
|
|
|
|
<title><acronym>GSSAPI</acronym> Session Encryption</title>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
If <productname>PostgreSQL</productname> was built with
|
|
|
|
<acronym>GSSAPI</acronym> support, frontend/backend communications
|
|
|
|
can be encrypted using <acronym>GSSAPI</acronym>. This provides
|
|
|
|
communication security in environments where attackers might be
|
|
|
|
able to capture the session traffic. For more information on
|
|
|
|
encrypting <productname>PostgreSQL</productname> sessions with
|
|
|
|
<acronym>GSSAPI</acronym>, see <xref linkend="gssapi-enc"/>.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
To initiate a <acronym>GSSAPI</acronym>-encrypted connection, the
|
|
|
|
frontend initially sends a GSSENCRequest message rather than a
|
|
|
|
StartupMessage. The server then responds with a single byte
|
|
|
|
containing <literal>G</literal> or <literal>N</literal>, indicating that it
|
|
|
|
is willing or unwilling to perform <acronym>GSSAPI</acronym> encryption,
|
|
|
|
respectively. The frontend might close the connection at this point
|
|
|
|
if it is dissatisfied with the response. To continue after
|
2020-12-01 13:36:30 +01:00
|
|
|
<literal>G</literal>, using the GSSAPI C bindings as discussed in
|
|
|
|
<ulink url="https://tools.ietf.org/html/rfc2744">RFC 2744</ulink>
|
2019-05-17 01:22:28 +02:00
|
|
|
or equivalent, perform a <acronym>GSSAPI</acronym> initialization by
|
2019-04-20 03:22:22 +02:00
|
|
|
calling <function>gss_init_sec_context()</function> in a loop and sending
|
|
|
|
the result to the server, starting with an empty input and then with each
|
|
|
|
result from the server, until it returns no output. When sending the
|
|
|
|
results of <function>gss_init_sec_context()</function> to the server,
|
|
|
|
prepend the length of the message as a four byte integer in network byte
|
2020-12-28 23:44:17 +01:00
|
|
|
order.
|
|
|
|
To continue after
|
2019-04-20 03:22:22 +02:00
|
|
|
<literal>N</literal>, send the usual StartupMessage and proceed without
|
|
|
|
encryption.
|
2020-12-28 23:44:17 +01:00
|
|
|
(Alternatively, it is permissible to issue an SSLRequest message
|
|
|
|
after an <literal>N</literal> response to try to
|
|
|
|
use <acronym>SSL</acronym> encryption instead
|
|
|
|
of <acronym>GSSAPI</acronym>.)
|
2019-04-20 03:22:22 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The frontend should also be prepared to handle an ErrorMessage
|
|
|
|
response to GSSENCRequest from the server. This would only occur if
|
|
|
|
the server predates the addition of <acronym>GSSAPI</acronym> encryption
|
|
|
|
support to <productname>PostgreSQL</productname>. In this case the
|
|
|
|
connection must be closed, but the frontend might choose to open a fresh
|
|
|
|
connection and proceed without requesting <acronym>GSSAPI</acronym>
|
2020-12-28 23:44:17 +01:00
|
|
|
encryption.
|
2019-04-20 03:22:22 +02:00
|
|
|
</para>
|
|
|
|
|
2021-11-08 17:14:56 +01:00
|
|
|
<para>
|
|
|
|
When <acronym>GSSAPI</acronym> encryption can be performed, the server
|
|
|
|
is expected to send only the single <literal>G</literal> byte and then
|
|
|
|
wait for the frontend to initiate a <acronym>GSSAPI</acronym> handshake.
|
|
|
|
If additional bytes are available to read at this point, it likely
|
|
|
|
means that a man-in-the-middle is attempting to perform a
|
|
|
|
buffer-stuffing attack
|
|
|
|
(<ulink url="https://www.postgresql.org/support/security/CVE-2021-23222/">CVE-2021-23222</ulink>).
|
|
|
|
Frontends should be coded either to read exactly one byte from the
|
|
|
|
socket before turning the socket over to their GSSAPI library, or to
|
|
|
|
treat it as a protocol violation if they find they have read additional
|
|
|
|
bytes.
|
|
|
|
</para>
|
|
|
|
|
2019-04-20 03:22:22 +02:00
|
|
|
<para>
|
|
|
|
An initial GSSENCRequest can also be used in a connection that is being
|
|
|
|
opened to send a CancelRequest message.
|
|
|
|
</para>
|
|
|
|
|
2020-12-28 23:44:17 +01:00
|
|
|
<para>
|
|
|
|
Once <acronym>GSSAPI</acronym> encryption has been successfully
|
|
|
|
established, use <function>gss_wrap()</function> to
|
|
|
|
encrypt the usual StartupMessage and all subsequent data, prepending the
|
|
|
|
length of the result from <function>gss_wrap()</function> as a four byte
|
|
|
|
integer in network byte order to the actual encrypted payload. Note that
|
|
|
|
the server will only accept encrypted packets from the client which are less
|
|
|
|
than 16kB; <function>gss_wrap_size_limit()</function> should be used by the
|
|
|
|
client to determine the size of the unencrypted message which will fit
|
|
|
|
within this limit and larger messages should be broken up into multiple
|
|
|
|
<function>gss_wrap()</function> calls. Typical segments are 8kB of
|
|
|
|
unencrypted data, resulting in encrypted packets of slightly larger than 8kB
|
|
|
|
but well within the 16kB maximum. The server can be expected to not send
|
|
|
|
encrypted packets of larger than 16kB to the client.
|
|
|
|
</para>
|
|
|
|
|
2019-04-20 03:22:22 +02:00
|
|
|
<para>
|
|
|
|
While the protocol itself does not provide a way for the server to
|
|
|
|
force <acronym>GSSAPI</acronym> encryption, the administrator can
|
|
|
|
configure the server to reject unencrypted sessions as a byproduct
|
|
|
|
of authentication checking.
|
|
|
|
</para>
|
|
|
|
</sect2>
|
2001-06-23 01:27:48 +02:00
|
|
|
</sect1>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
<sect1 id="sasl-authentication">
|
|
|
|
<title>SASL Authentication</title>
|
|
|
|
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
<firstterm>SASL</firstterm> is a framework for authentication in connection-oriented
|
2017-11-18 16:07:57 +01:00
|
|
|
protocols. At the moment, <productname>PostgreSQL</productname> implements two SASL
|
|
|
|
authentication mechanisms, SCRAM-SHA-256 and SCRAM-SHA-256-PLUS. More
|
|
|
|
might be added in the future. The below steps illustrate how SASL
|
|
|
|
authentication is performed in general, while the next subsection gives
|
|
|
|
more details on SCRAM-SHA-256 and SCRAM-SHA-256-PLUS.
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<procedure>
|
|
|
|
<title>SASL Authentication Message Flow</title>
|
|
|
|
|
|
|
|
<step id="sasl-auth-begin">
|
2017-06-12 15:51:18 +02:00
|
|
|
<para>
|
2017-04-19 20:42:27 +02:00
|
|
|
To begin a SASL authentication exchange, the server sends an
|
|
|
|
AuthenticationSASL message. It includes a list of SASL authentication
|
|
|
|
mechanisms that the server can accept, in the server's preferred order.
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</para>
|
|
|
|
</step>
|
|
|
|
|
|
|
|
<step id="sasl-auth-initial-response">
|
|
|
|
<para>
|
|
|
|
The client selects one of the supported mechanisms from the list, and sends
|
|
|
|
a SASLInitialResponse message to the server. The message includes the name
|
|
|
|
of the selected mechanism, and an optional Initial Client Response, if the
|
|
|
|
selected mechanism uses that.
|
|
|
|
</para>
|
|
|
|
</step>
|
|
|
|
|
|
|
|
<step id="sasl-auth-continue">
|
|
|
|
<para>
|
|
|
|
One or more server-challenge and client-response message will follow. Each
|
|
|
|
server-challenge is sent in an AuthenticationSASLContinue message, followed
|
2021-06-11 03:38:04 +02:00
|
|
|
by a response from client in a SASLResponse message. The particulars of
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
the messages are mechanism specific.
|
|
|
|
</para>
|
|
|
|
</step>
|
|
|
|
|
|
|
|
<step id="sasl-auth-end">
|
|
|
|
<para>
|
|
|
|
Finally, when the authentication exchange is completed successfully, the
|
|
|
|
server sends an AuthenticationSASLFinal message, followed
|
|
|
|
immediately by an AuthenticationOk message. The AuthenticationSASLFinal
|
|
|
|
contains additional server-to-client data, whose content is particular to the
|
|
|
|
selected authentication mechanism. If the authentication mechanism doesn't
|
|
|
|
use additional data that's sent at completion, the AuthenticationSASLFinal
|
|
|
|
message is not sent.
|
|
|
|
</para>
|
|
|
|
</step>
|
|
|
|
</procedure>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
On error, the server can abort the authentication at any stage, and send an
|
|
|
|
ErrorMessage.
|
|
|
|
</para>
|
|
|
|
|
2018-01-30 22:50:30 +01:00
|
|
|
<sect2 id="sasl-scram-sha-256">
|
2019-09-08 10:26:35 +02:00
|
|
|
<title>SCRAM-SHA-256 Authentication</title>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
|
|
|
|
<para>
|
2017-11-18 16:07:57 +01:00
|
|
|
The implemented SASL mechanisms at the moment
|
|
|
|
are <literal>SCRAM-SHA-256</literal> and its variant with channel
|
|
|
|
binding <literal>SCRAM-SHA-256-PLUS</literal>. They are described in
|
2020-12-01 13:36:30 +01:00
|
|
|
detail in <ulink url="https://tools.ietf.org/html/rfc7677">RFC 7677</ulink>
|
|
|
|
and <ulink url="https://tools.ietf.org/html/rfc5802">RFC 5802</ulink>.
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2017-08-23 18:01:43 +02:00
|
|
|
When SCRAM-SHA-256 is used in PostgreSQL, the server will ignore the user name
|
2017-10-09 03:44:17 +02:00
|
|
|
that the client sends in the <structname>client-first-message</structname>. The user name
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
that was already sent in the startup message is used instead.
|
2017-10-09 03:44:17 +02:00
|
|
|
<productname>PostgreSQL</productname> supports multiple character encodings, while SCRAM
|
2017-08-23 18:01:43 +02:00
|
|
|
dictates UTF-8 to be used for the user name, so it might be impossible to
|
2017-09-13 16:10:34 +02:00
|
|
|
represent the PostgreSQL user name in UTF-8.
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The SCRAM specification dictates that the password is also in UTF-8, and is
|
2017-10-09 03:44:17 +02:00
|
|
|
processed with the <firstterm>SASLprep</firstterm> algorithm.
|
|
|
|
<productname>PostgreSQL</productname>, however, does not require UTF-8 to be used for
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
the password. When a user's password is set, it is processed with SASLprep
|
|
|
|
as if it was in UTF-8, regardless of the actual encoding used. However, if
|
|
|
|
it is not a legal UTF-8 byte sequence, or it contains UTF-8 byte sequences
|
|
|
|
that are prohibited by the SASLprep algorithm, the raw password will be used
|
|
|
|
without SASLprep processing, instead of throwing an error. This allows the
|
|
|
|
password to be normalized when it is in UTF-8, but still allows a non-UTF-8
|
|
|
|
password to be used, and doesn't require the system to know which encoding
|
|
|
|
the password is in.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2017-11-18 16:07:57 +01:00
|
|
|
<firstterm>Channel binding</firstterm> is supported in PostgreSQL builds with
|
2018-01-04 21:18:39 +01:00
|
|
|
SSL support. The SASL mechanism name for SCRAM with channel binding is
|
Remove support for tls-unique channel binding.
There are some problems with the tls-unique channel binding type. It's not
supported by all SSL libraries, and strictly speaking it's not defined for
TLS 1.3 at all, even though at least in OpenSSL, the functions used for it
still seem to work with TLS 1.3 connections. And since we had no
mechanism to negotiate what channel binding type to use, there would be
awkward interoperability issues if a server only supported some channel
binding types. tls-server-end-point seems feasible to support with any SSL
library, so let's just stick to that.
This removes the scram_channel_binding libpq option altogether, since there
is now only one supported channel binding type.
This also removes all the channel binding tests from the SSL test suite.
They were really just testing the scram_channel_binding option, which
is now gone. Channel binding is used if both client and server support it,
so it is used in the existing tests. It would be good to have some tests
specifically for channel binding, to make sure it really is used, and the
different combinations of a client and a server that support or doesn't
support it. The current set of settings we have make it hard to write such
tests, but I did test those things manually, by disabling
HAVE_BE_TLS_GET_CERTIFICATE_HASH and/or
HAVE_PGTLS_GET_PEER_CERTIFICATE_HASH.
I also removed the SCRAM_CHANNEL_BINDING_TLS_END_POINT constant. This is a
matter of taste, but IMO it's more readable to just use the
"tls-server-end-point" string.
Refactor the checks on whether the SSL library supports the functions
needed for tls-server-end-point channel binding. Now the server won't
advertise, and the client won't choose, the SCRAM-SHA-256-PLUS variant, if
compiled with an OpenSSL version too old to support it.
In the passing, add some sanity checks to check that the chosen SASL
mechanism, SCRAM-SHA-256 or SCRAM-SHA-256-PLUS, matches whether the SCRAM
exchange used channel binding or not. For example, if the client selects
the non-channel-binding variant SCRAM-SHA-256, but in the SCRAM message
uses channel binding anyway. It's harmless from a security point of view,
I believe, and I'm not sure if there are some other conditions that would
cause the connection to fail, but it seems better to be strict about these
things and check explicitly.
Discussion: https://www.postgresql.org/message-id/ec787074-2305-c6f4-86aa-6902f98485a4%40iki.fi
2018-08-05 12:44:21 +02:00
|
|
|
<literal>SCRAM-SHA-256-PLUS</literal>. The channel binding type used by
|
|
|
|
PostgreSQL is <literal>tls-server-end-point</literal>.
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</para>
|
|
|
|
|
2018-05-15 02:45:35 +02:00
|
|
|
<para>
|
|
|
|
In <acronym>SCRAM</acronym> without channel binding, the server chooses
|
|
|
|
a random number that is transmitted to the client to be mixed with the
|
|
|
|
user-supplied password in the transmitted password hash. While this
|
|
|
|
prevents the password hash from being successfully retransmitted in
|
|
|
|
a later session, it does not prevent a fake server between the real
|
2019-04-08 22:27:35 +02:00
|
|
|
server and client from passing through the server's random value
|
2018-05-15 02:45:35 +02:00
|
|
|
and successfully authenticating.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
<acronym>SCRAM</acronym> with channel binding prevents such
|
Remove support for tls-unique channel binding.
There are some problems with the tls-unique channel binding type. It's not
supported by all SSL libraries, and strictly speaking it's not defined for
TLS 1.3 at all, even though at least in OpenSSL, the functions used for it
still seem to work with TLS 1.3 connections. And since we had no
mechanism to negotiate what channel binding type to use, there would be
awkward interoperability issues if a server only supported some channel
binding types. tls-server-end-point seems feasible to support with any SSL
library, so let's just stick to that.
This removes the scram_channel_binding libpq option altogether, since there
is now only one supported channel binding type.
This also removes all the channel binding tests from the SSL test suite.
They were really just testing the scram_channel_binding option, which
is now gone. Channel binding is used if both client and server support it,
so it is used in the existing tests. It would be good to have some tests
specifically for channel binding, to make sure it really is used, and the
different combinations of a client and a server that support or doesn't
support it. The current set of settings we have make it hard to write such
tests, but I did test those things manually, by disabling
HAVE_BE_TLS_GET_CERTIFICATE_HASH and/or
HAVE_PGTLS_GET_PEER_CERTIFICATE_HASH.
I also removed the SCRAM_CHANNEL_BINDING_TLS_END_POINT constant. This is a
matter of taste, but IMO it's more readable to just use the
"tls-server-end-point" string.
Refactor the checks on whether the SSL library supports the functions
needed for tls-server-end-point channel binding. Now the server won't
advertise, and the client won't choose, the SCRAM-SHA-256-PLUS variant, if
compiled with an OpenSSL version too old to support it.
In the passing, add some sanity checks to check that the chosen SASL
mechanism, SCRAM-SHA-256 or SCRAM-SHA-256-PLUS, matches whether the SCRAM
exchange used channel binding or not. For example, if the client selects
the non-channel-binding variant SCRAM-SHA-256, but in the SCRAM message
uses channel binding anyway. It's harmless from a security point of view,
I believe, and I'm not sure if there are some other conditions that would
cause the connection to fail, but it seems better to be strict about these
things and check explicitly.
Discussion: https://www.postgresql.org/message-id/ec787074-2305-c6f4-86aa-6902f98485a4%40iki.fi
2018-08-05 12:44:21 +02:00
|
|
|
man-in-the-middle attacks by mixing the signature of the server's
|
|
|
|
certificate into the transmitted password hash. While a fake server can
|
|
|
|
retransmit the real server's certificate, it doesn't have access to the
|
|
|
|
private key matching that certificate, and therefore cannot prove it is
|
|
|
|
the owner, causing SSL connection failure.
|
2018-05-15 02:45:35 +02:00
|
|
|
</para>
|
|
|
|
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
<procedure>
|
|
|
|
<title>Example</title>
|
|
|
|
<step id="scram-begin">
|
|
|
|
<para>
|
|
|
|
The server sends an AuthenticationSASL message. It includes a list of
|
|
|
|
SASL authentication mechanisms that the server can accept.
|
2017-11-18 16:07:57 +01:00
|
|
|
This will be <literal>SCRAM-SHA-256-PLUS</literal>
|
|
|
|
and <literal>SCRAM-SHA-256</literal> if the server is built with SSL
|
|
|
|
support, or else just the latter.
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</para>
|
|
|
|
</step>
|
|
|
|
<step id="scram-client-first">
|
|
|
|
<para>
|
|
|
|
The client responds by sending a SASLInitialResponse message, which
|
2017-11-18 16:07:57 +01:00
|
|
|
indicates the chosen mechanism, <literal>SCRAM-SHA-256</literal> or
|
|
|
|
<literal>SCRAM-SHA-256-PLUS</literal>. (A client is free to choose either
|
|
|
|
mechanism, but for better security it should choose the channel-binding
|
2018-01-04 21:18:39 +01:00
|
|
|
variant if it can support it.) In the Initial Client response field, the
|
|
|
|
message contains the SCRAM <structname>client-first-message</structname>.
|
|
|
|
The <structname>client-first-message</structname> also contains the channel
|
|
|
|
binding type chosen by the client.
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</para>
|
|
|
|
</step>
|
|
|
|
<step id="scram-server-first">
|
|
|
|
<para>
|
|
|
|
Server sends an AuthenticationSASLContinue message, with a SCRAM
|
2018-11-23 11:41:27 +01:00
|
|
|
<structname>server-first-message</structname> as the content.
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</para>
|
|
|
|
</step>
|
|
|
|
<step id="scram-client-final">
|
|
|
|
<para>
|
|
|
|
Client sends a SASLResponse message, with SCRAM
|
2017-10-09 03:44:17 +02:00
|
|
|
<structname>client-final-message</structname> as the content.
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</para>
|
|
|
|
</step>
|
|
|
|
<step id="scram-server-final">
|
|
|
|
<para>
|
|
|
|
Server sends an AuthenticationSASLFinal message, with the SCRAM
|
2017-10-09 03:44:17 +02:00
|
|
|
<structname>server-final-message</structname>, followed immediately by
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
an AuthenticationOk message.
|
|
|
|
</para>
|
|
|
|
</step>
|
|
|
|
</procedure>
|
|
|
|
</sect2>
|
2017-04-13 19:12:58 +02:00
|
|
|
</sect1>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
|
2010-06-04 00:17:32 +02:00
|
|
|
<sect1 id="protocol-replication">
|
|
|
|
<title>Streaming Replication Protocol</title>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
To initiate streaming replication, the frontend sends the
|
2018-03-07 03:00:10 +01:00
|
|
|
<literal>replication</literal> parameter in the startup message. A Boolean
|
|
|
|
value of <literal>true</literal> (or <literal>on</literal>,
|
|
|
|
<literal>yes</literal>, <literal>1</literal>) tells the backend to go into
|
|
|
|
physical replication walsender mode, wherein a small set of replication
|
|
|
|
commands, shown below, can be issued instead of SQL statements.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Passing <literal>database</literal> as the value for the
|
|
|
|
<literal>replication</literal> parameter instructs the backend to go into
|
|
|
|
logical replication walsender mode, connecting to the database specified in
|
|
|
|
the <literal>dbname</literal> parameter. In logical replication walsender
|
|
|
|
mode, the replication commands shown below as well as normal SQL commands can
|
|
|
|
be issued.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
In either physical replication or logical replication walsender mode, only the
|
|
|
|
simple query protocol can be used.
|
2014-03-18 18:20:01 +01:00
|
|
|
</para>
|
2018-03-07 03:00:10 +01:00
|
|
|
|
2014-03-18 18:20:01 +01:00
|
|
|
<para>
|
|
|
|
For the purpose of testing replication commands, you can make a replication
|
2020-08-22 15:26:10 +02:00
|
|
|
connection via <application>psql</application> or any other
|
|
|
|
<application>libpq</application>-using tool with a connection string including
|
|
|
|
the <literal>replication</literal> option,
|
2014-03-18 18:20:01 +01:00
|
|
|
e.g.:
|
2014-07-08 17:39:07 +02:00
|
|
|
<programlisting>
|
|
|
|
psql "dbname=postgres replication=database" -c "IDENTIFY_SYSTEM;"
|
|
|
|
</programlisting>
|
2015-10-22 04:31:56 +02:00
|
|
|
However, it is often more useful to use
|
2017-11-23 15:39:47 +01:00
|
|
|
<xref linkend="app-pgreceivewal"/> (for physical replication) or
|
|
|
|
<xref linkend="app-pgrecvlogical"/> (for logical replication).
|
2014-03-18 18:20:01 +01:00
|
|
|
</para>
|
2010-06-04 00:17:32 +02:00
|
|
|
|
2014-03-18 18:20:01 +01:00
|
|
|
<para>
|
2018-03-07 03:00:10 +01:00
|
|
|
Replication commands are logged in the server log when
|
|
|
|
<xref linkend="guc-log-replication-commands"/> is enabled.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The commands accepted in replication mode are:
|
2010-06-04 00:17:32 +02:00
|
|
|
<variablelist>
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-identify-system">
|
2015-10-22 04:31:56 +02:00
|
|
|
<term><literal>IDENTIFY_SYSTEM</literal>
|
2014-08-18 04:18:53 +02:00
|
|
|
<indexterm><primary>IDENTIFY_SYSTEM</primary></indexterm>
|
|
|
|
</term>
|
2010-06-04 00:17:32 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Requests the server to identify itself. Server replies with a result
|
2014-03-10 18:50:28 +01:00
|
|
|
set of a single row, containing four fields:
|
2010-06-04 00:17:32 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2015-10-22 04:31:56 +02:00
|
|
|
<literal>systemid</literal> (<type>text</type>)
|
2010-06-04 00:17:32 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The unique system identifier identifying the cluster. This
|
|
|
|
can be used to check that the base backup used to initialize the
|
2010-06-07 04:01:09 +02:00
|
|
|
standby came from the same cluster.
|
2010-06-04 00:17:32 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2015-10-22 04:31:56 +02:00
|
|
|
<literal>timeline</literal> (<type>int4</type>)
|
2010-06-04 00:17:32 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2015-10-22 04:31:56 +02:00
|
|
|
Current timeline ID. Also useful to check that the standby is
|
2020-06-15 19:12:58 +02:00
|
|
|
consistent with the primary.
|
2010-06-04 00:17:32 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2011-02-03 13:46:23 +01:00
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2015-10-22 04:31:56 +02:00
|
|
|
<literal>xlogpos</literal> (<type>text</type>)
|
2011-02-03 13:46:23 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-02-13 18:30:46 +01:00
|
|
|
Current WAL flush location. Useful to get a known location in the
|
2017-05-12 17:49:56 +02:00
|
|
|
write-ahead log where streaming can start.
|
2011-02-03 13:46:23 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2014-03-10 18:50:28 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2015-10-22 04:31:56 +02:00
|
|
|
<literal>dbname</literal> (<type>text</type>)
|
2014-03-10 18:50:28 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2015-10-22 04:31:56 +02:00
|
|
|
Database connected to or null.
|
2014-03-10 18:50:28 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2010-06-04 00:17:32 +02:00
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-show">
|
2017-01-24 22:59:18 +01:00
|
|
|
<term><literal>SHOW</literal> <replaceable class="parameter">name</replaceable>
|
|
|
|
<indexterm><primary>SHOW</primary></indexterm>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Requests the server to send the current setting of a run-time parameter.
|
2017-11-23 15:39:47 +01:00
|
|
|
This is similar to the SQL command <xref linkend="sql-show"/>.
|
2017-01-24 22:59:18 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><replaceable class="parameter">name</replaceable></term>
|
2017-01-24 22:59:18 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The name of a run-time parameter. Available parameters are documented
|
2017-11-23 15:39:47 +01:00
|
|
|
in <xref linkend="runtime-config"/>.
|
2017-01-24 22:59:18 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-timeline-history">
|
2015-10-22 04:31:56 +02:00
|
|
|
<term><literal>TIMELINE_HISTORY</literal> <replaceable class="parameter">tli</replaceable>
|
2014-08-18 04:18:53 +02:00
|
|
|
<indexterm><primary>TIMELINE_HISTORY</primary></indexterm>
|
|
|
|
</term>
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Requests the server to send over the timeline history file for timeline
|
|
|
|
<replaceable class="parameter">tli</replaceable>. Server replies with a
|
2020-11-12 20:08:59 +01:00
|
|
|
result set of a single row, containing two fields. While the fields
|
|
|
|
are labeled as <type>text</type>, they effectively return raw bytes,
|
|
|
|
with no encoding conversion:
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2015-10-22 04:31:56 +02:00
|
|
|
<literal>filename</literal> (<type>text</type>)
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
File name of the timeline history file, e.g., <filename>00000002.history</filename>.
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2020-11-12 20:08:59 +01:00
|
|
|
<literal>content</literal> (<type>text</type>)
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Contents of the timeline history file.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-create-replication-slot" xreflabel="CREATE_REPLICATION_SLOT">
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
<term><literal>CREATE_REPLICATION_SLOT</literal> <replaceable class="parameter">slot_name</replaceable> [ <literal>TEMPORARY</literal> ] { <literal>PHYSICAL</literal> | <literal>LOGICAL</literal> } [ ( <replaceable class="parameter">option</replaceable> [, ...] ) ]
|
2014-08-18 04:18:53 +02:00
|
|
|
<indexterm><primary>CREATE_REPLICATION_SLOT</primary></indexterm>
|
|
|
|
</term>
|
2014-02-01 04:45:17 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2014-05-31 15:58:04 +02:00
|
|
|
Create a physical or logical replication
|
2017-11-23 15:39:47 +01:00
|
|
|
slot. See <xref linkend="streaming-replication-slots"/> for more about
|
2014-02-01 04:45:17 +01:00
|
|
|
replication slots.
|
|
|
|
</para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><replaceable class="parameter">slot_name</replaceable></term>
|
2014-02-01 04:45:17 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The name of the slot to create. Must be a valid replication slot
|
2017-11-23 15:39:47 +01:00
|
|
|
name (see <xref linkend="streaming-replication-slots-manipulation"/>).
|
2014-02-01 04:45:17 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2014-05-31 15:58:04 +02:00
|
|
|
|
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><replaceable class="parameter">output_plugin</replaceable></term>
|
2014-05-31 15:58:04 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The name of the output plugin used for logical decoding
|
2017-11-23 15:39:47 +01:00
|
|
|
(see <xref linkend="logicaldecoding-output-plugin"/>).
|
2014-05-31 15:58:04 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2015-09-06 13:17:23 +02:00
|
|
|
|
2016-12-08 18:00:00 +01:00
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><literal>TEMPORARY</literal></term>
|
2016-12-08 18:00:00 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Specify that this replication slot is a temporary one. Temporary
|
|
|
|
slots are not saved to disk and are automatically dropped on error
|
|
|
|
or when the session has finished.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
<para>The following options are supported:</para>
|
2016-12-08 18:00:00 +01:00
|
|
|
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
<variablelist>
|
2021-06-30 05:15:47 +02:00
|
|
|
<varlistentry>
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
<term><literal>TWO_PHASE [ <replaceable class="parameter">boolean</replaceable> ]</literal></term>
|
2021-06-30 05:15:47 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
If true, this logical replication slot supports decoding of two-phase
|
2021-06-30 05:15:47 +02:00
|
|
|
transactions. With this option, two-phase commands like
|
|
|
|
<literal>PREPARE TRANSACTION</literal>, <literal>COMMIT PREPARED</literal>
|
|
|
|
and <literal>ROLLBACK PREPARED</literal> are decoded and transmitted.
|
|
|
|
The transaction will be decoded and transmitted at
|
|
|
|
<literal>PREPARE TRANSACTION</literal> time.
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
The default is false.
|
2021-06-30 05:15:47 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2015-09-06 13:17:23 +02:00
|
|
|
<varlistentry>
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
<term><literal>RESERVE_WAL [ <replaceable class="parameter">boolean</replaceable> ]</literal></term>
|
2015-09-06 13:17:23 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
If true, this physical replication slot reserves <acronym>WAL</acronym>
|
2017-10-09 03:44:17 +02:00
|
|
|
immediately. Otherwise, <acronym>WAL</acronym> is only reserved upon
|
2015-09-06 13:17:23 +02:00
|
|
|
connection from a streaming replication client.
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
The default is false.
|
2015-09-06 13:17:23 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2017-03-14 22:13:56 +01:00
|
|
|
|
|
|
|
<varlistentry>
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
<term><literal>SNAPSHOT { 'export' | 'use' | 'nothing' }</literal></term>
|
2017-03-14 22:13:56 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Decides what to do with the snapshot created during logical slot
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
initialization. <literal>'export'</literal>, which is the default,
|
2017-03-14 22:13:56 +01:00
|
|
|
will export the snapshot for use in other sessions. This option can't
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
be used inside a transaction. <literal>'use'</literal> will use the
|
2017-03-23 13:36:36 +01:00
|
|
|
snapshot for the current transaction executing the command. This
|
|
|
|
option must be used in a transaction, and
|
|
|
|
<literal>CREATE_REPLICATION_SLOT</literal> must be the first command
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
run in that transaction. Finally, <literal>'nothing'</literal> will
|
2017-03-14 22:13:56 +01:00
|
|
|
just use the snapshot for logical decoding as normal but won't do
|
|
|
|
anything else with it.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2014-02-01 04:45:17 +01:00
|
|
|
</variablelist>
|
2017-02-02 22:04:59 +01:00
|
|
|
|
|
|
|
<para>
|
|
|
|
In response to this command, the server will send a one-row result set
|
|
|
|
containing the following fields:
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term><literal>slot_name</literal> (<type>text</type>)</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The name of the newly-created replication slot.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term><literal>consistent_point</literal> (<type>text</type>)</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-05-12 19:51:27 +02:00
|
|
|
The WAL location at which the slot became consistent. This is the
|
2017-02-02 22:04:59 +01:00
|
|
|
earliest location from which streaming can start on this replication
|
|
|
|
slot.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term><literal>snapshot_name</literal> (<type>text</type>)</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The identifier of the snapshot exported by the command. The
|
|
|
|
snapshot is valid until a new command is executed on this connection
|
|
|
|
or the replication connection is closed. Null if the created slot
|
|
|
|
is physical.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term><literal>output_plugin</literal> (<type>text</type>)</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The name of the output plugin used by the newly-created replication
|
|
|
|
slot. Null if the created slot is physical.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
2014-02-01 04:45:17 +01:00
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-create-replication-slot-legacy">
|
Flexible options for CREATE_REPLICATION_SLOT.
Like BASE_BACKUP, CREATE_REPLICATION_SLOT has historically used a
hard-coded syntax. To improve future extensibility, adopt a flexible
options syntax here, too.
In the new syntax, instead of three mutually exclusive options
EXPORT_SNAPSHOT, USE_SNAPSHOT, and NOEXPORT_SNAPSHOT, there is now a single
SNAPSHOT option with three possible values: 'export', 'use', and 'nothing'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, makes pg_receivewal,
pg_recvlogical, and walreceiver processes use it.
Patch by me, reviewed by Fabien Coelho, Sergei Kornilov, and
Fujii Masao.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 18:52:49 +02:00
|
|
|
<term><literal>CREATE_REPLICATION_SLOT</literal> <replaceable class="parameter">slot_name</replaceable> [ <literal>TEMPORARY</literal> ] { <literal>PHYSICAL</literal> [ <literal>RESERVE_WAL</literal> ] | <literal>LOGICAL</literal> <replaceable class="parameter">output_plugin</replaceable> [ <literal>EXPORT_SNAPSHOT</literal> | <literal>NOEXPORT_SNAPSHOT</literal> | <literal>USE_SNAPSHOT</literal> | <literal>TWO_PHASE</literal> ] }
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
For compatibility with older releases, this alternative syntax for
|
|
|
|
the <literal>CREATE_REPLICATION_SLOT</literal> command is still supported.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-read-replication-slot">
|
2021-10-25 00:40:42 +02:00
|
|
|
<term><literal>READ_REPLICATION_SLOT</literal> <replaceable class="parameter">slot_name</replaceable>
|
|
|
|
<indexterm><primary>READ_REPLICATION_SLOT</primary></indexterm>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2022-01-25 01:40:04 +01:00
|
|
|
Read some information associated with a replication slot. Returns a tuple
|
2021-10-25 00:40:42 +02:00
|
|
|
with <literal>NULL</literal> values if the replication slot does not
|
|
|
|
exist. This command is currently only supported for physical replication
|
|
|
|
slots.
|
|
|
|
</para>
|
|
|
|
<para>
|
|
|
|
In response to this command, the server will return a one-row result set,
|
|
|
|
containing the following fields:
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term><literal>slot_type</literal> (<type>text</type>)</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The replication slot's type, either <literal>physical</literal> or
|
|
|
|
<literal>NULL</literal>.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term><literal>restart_lsn</literal> (<type>text</type>)</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The replication slot's <literal>restart_lsn</literal>.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term><literal>restart_tli</literal> (<type>int8</type>)</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2022-01-25 01:40:04 +01:00
|
|
|
The timeline ID associated with <literal>restart_lsn</literal>,
|
2021-10-25 00:40:42 +02:00
|
|
|
following the current timeline history.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-start-replication">
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><literal>START_REPLICATION</literal> [ <literal>SLOT</literal> <replaceable class="parameter">slot_name</replaceable> ] [ <literal>PHYSICAL</literal> ] <replaceable class="parameter">XXX/XXX</replaceable> [ <literal>TIMELINE</literal> <replaceable class="parameter">tli</replaceable> ]
|
2014-08-18 04:18:53 +02:00
|
|
|
<indexterm><primary>START_REPLICATION</primary></indexterm>
|
|
|
|
</term>
|
2010-06-04 00:17:32 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Instructs server to start streaming WAL, starting at
|
2017-10-09 03:44:17 +02:00
|
|
|
WAL location <replaceable class="parameter">XXX/XXX</replaceable>.
|
2014-03-24 11:23:32 +01:00
|
|
|
If <literal>TIMELINE</literal> option is specified,
|
2017-10-09 03:44:17 +02:00
|
|
|
streaming starts on timeline <replaceable class="parameter">tli</replaceable>;
|
2014-02-01 04:45:17 +01:00
|
|
|
otherwise, the server's current timeline is selected. The server can
|
2015-10-22 04:31:56 +02:00
|
|
|
reply with an error, for example if the requested section of WAL has already
|
2022-01-25 01:40:04 +01:00
|
|
|
been recycled. On success, the server responds with a CopyBothResponse
|
2014-02-01 04:45:17 +01:00
|
|
|
message, and then starts to stream WAL to the frontend.
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
|
2014-11-04 22:10:58 +01:00
|
|
|
<para>
|
|
|
|
If a slot's name is provided
|
2017-10-09 03:44:17 +02:00
|
|
|
via <replaceable class="parameter">slot_name</replaceable>, it will be updated
|
2014-11-04 22:10:58 +01:00
|
|
|
as replication progresses so that the server knows which WAL segments,
|
2017-10-09 03:44:17 +02:00
|
|
|
and if <varname>hot_standby_feedback</varname> is on which transactions,
|
2014-11-04 22:10:58 +01:00
|
|
|
are still needed by the standby.
|
|
|
|
</para>
|
|
|
|
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
<para>
|
2015-10-22 04:31:56 +02:00
|
|
|
If the client requests a timeline that's not the latest but is part of
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
the history of the server, the server will stream all the WAL on that
|
2015-10-22 04:31:56 +02:00
|
|
|
timeline starting from the requested start point up to the point where
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
the server switched to another timeline. If the client requests
|
2021-01-16 23:40:12 +01:00
|
|
|
streaming at exactly the end of an old timeline, the server skips COPY
|
|
|
|
mode entirely.
|
Allow a streaming replication standby to follow a timeline switch.
Before this patch, streaming replication would refuse to start replicating
if the timeline in the primary doesn't exactly match the standby. The
situation where it doesn't match is when you have a master, and two
standbys, and you promote one of the standbys to become new master.
Promoting bumps up the timeline ID, and after that bump, the other standby
would refuse to continue.
There's significantly more timeline related logic in streaming replication
now. First of all, when a standby connects to primary, it will ask the
primary for any timeline history files that are missing from the standby.
The missing files are sent using a new replication command TIMELINE_HISTORY,
and stored in standby's pg_xlog directory. Using the timeline history files,
the standby can follow the latest timeline present in the primary
(recovery_target_timeline='latest'), just as it can follow new timelines
appearing in an archive directory.
START_REPLICATION now takes a TIMELINE parameter, to specify exactly which
timeline to stream WAL from. This allows the standby to request the primary
to send over WAL that precedes the promotion. The replication protocol is
changed slightly (in a backwards-compatible way although there's little hope
of streaming replication working across major versions anyway), to allow
replication to stop when the end of timeline reached, putting the walsender
back into accepting a replication command.
Many thanks to Amit Kapila for testing and reviewing various versions of
this patch.
2012-12-13 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
After streaming all the WAL on a timeline that is not the latest one,
|
|
|
|
the server will end streaming by exiting the COPY mode. When the client
|
Fix walsender failure at promotion.
If a standby server has a cascading standby server connected to it, it's
possible that WAL has already been sent up to the next WAL page boundary,
splitting a WAL record in the middle, when the first standby server is
promoted. Don't throw an assertion failure or error in walsender if that
happens.
Also, fix a variant of the same bug in pg_receivexlog: if it had already
received WAL on previous timeline up to a segment boundary, when the
upstream standby server is promoted so that the timeline switch record falls
on the previous segment, pg_receivexlog would miss the segment containing
the timeline switch. To fix that, have walsender send the position of the
timeline switch at end-of-streaming, in addition to the next timeline's ID.
It was previously assumed that the switch happened exactly where the
streaming stopped.
Note: this is an incompatible change in the streaming protocol. You might
get an error if you try to stream over timeline switches, if the client is
running 9.3beta1 and the server is more recent. It should be fine after a
reconnect, however.
Reported by Fujii Masao.
2013-05-08 19:10:17 +02:00
|
|
|
acknowledges this by also exiting COPY mode, the server sends a result
|
|
|
|
set with one row and two columns, indicating the next timeline in this
|
2015-10-22 04:31:56 +02:00
|
|
|
server's history. The first column is the next timeline's ID (type <type>int8</type>), and the
|
2017-05-12 19:51:27 +02:00
|
|
|
second column is the WAL location where the switch happened (type <type>text</type>). Usually,
|
Fix walsender failure at promotion.
If a standby server has a cascading standby server connected to it, it's
possible that WAL has already been sent up to the next WAL page boundary,
splitting a WAL record in the middle, when the first standby server is
promoted. Don't throw an assertion failure or error in walsender if that
happens.
Also, fix a variant of the same bug in pg_receivexlog: if it had already
received WAL on previous timeline up to a segment boundary, when the
upstream standby server is promoted so that the timeline switch record falls
on the previous segment, pg_receivexlog would miss the segment containing
the timeline switch. To fix that, have walsender send the position of the
timeline switch at end-of-streaming, in addition to the next timeline's ID.
It was previously assumed that the switch happened exactly where the
streaming stopped.
Note: this is an incompatible change in the streaming protocol. You might
get an error if you try to stream over timeline switches, if the client is
running 9.3beta1 and the server is more recent. It should be fine after a
reconnect, however.
Reported by Fujii Masao.
2013-05-08 19:10:17 +02:00
|
|
|
the switch position is the end of the WAL that was streamed, but there
|
|
|
|
are corner cases where the server can send some WAL from the old
|
|
|
|
timeline that it has not itself replayed before promoting. Finally, the
|
2020-10-15 01:12:26 +02:00
|
|
|
server sends two CommandComplete messages (one that ends the CopyData
|
|
|
|
and the other ends the <literal>START_REPLICATION</literal> itself), and
|
|
|
|
is ready to accept a new command.
|
2010-06-04 00:17:32 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
WAL data is sent as a series of CopyData messages. (This allows
|
|
|
|
other information to be intermixed; in particular the server can send
|
|
|
|
an ErrorResponse message if it encounters a failure after beginning
|
2012-11-07 17:59:12 +01:00
|
|
|
to stream.) The payload of each CopyData message from server to the
|
|
|
|
client contains a message of one of the following formats:
|
2010-06-04 00:17:32 +02:00
|
|
|
</para>
|
|
|
|
|
2014-02-01 04:45:17 +01:00
|
|
|
<para>
|
2010-06-04 00:17:32 +02:00
|
|
|
<variablelist>
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-xlogdata">
|
2010-06-04 00:17:32 +02:00
|
|
|
<term>
|
|
|
|
XLogData (B)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('w')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as WAL data.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int64
|
2010-06-04 00:17:32 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2012-11-07 17:59:12 +01:00
|
|
|
The starting point of the WAL data in this message.
|
2010-06-04 00:17:32 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int64
|
2010-06-04 00:17:32 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2012-11-07 17:59:12 +01:00
|
|
|
The current end of WAL on the server.
|
2010-06-04 00:17:32 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int64
|
2010-06-04 00:17:32 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2012-11-07 17:59:12 +01:00
|
|
|
The server's system clock at the time of transmission, as
|
|
|
|
microseconds since midnight on 2000-01-01.
|
2010-06-04 00:17:32 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
A section of the WAL data stream.
|
|
|
|
</para>
|
2012-11-07 17:59:12 +01:00
|
|
|
<para>
|
|
|
|
A single WAL record is never split across two XLogData messages.
|
|
|
|
When a WAL record crosses a WAL page boundary, and is therefore
|
|
|
|
already split using continuation records, it can be split at the page
|
|
|
|
boundary. In other words, the first main WAL record and its
|
|
|
|
continuation records can be sent in different XLogData messages.
|
|
|
|
</para>
|
2010-06-04 00:17:32 +02:00
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-primary-keepalive-message">
|
2011-12-31 14:30:26 +01:00
|
|
|
<term>
|
|
|
|
Primary keepalive message (B)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('k')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a sender keepalive.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int64
|
2011-12-31 14:30:26 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2012-11-07 17:59:12 +01:00
|
|
|
The current end of WAL on the server.
|
2011-12-31 14:30:26 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int64
|
2011-12-31 14:30:26 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2012-11-07 17:59:12 +01:00
|
|
|
The server's system clock at the time of transmission, as
|
|
|
|
microseconds since midnight on 2000-01-01.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
1 means that the client should reply to this message as soon as
|
|
|
|
possible, to avoid a timeout disconnect. 0 otherwise.
|
2011-12-31 14:30:26 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
|
2012-11-07 17:59:12 +01:00
|
|
|
<para>
|
|
|
|
The receiving process can send replies back to the sender at any time,
|
|
|
|
using one of the following message formats (also in the payload of a
|
|
|
|
CopyData message):
|
|
|
|
</para>
|
|
|
|
|
2011-02-10 20:00:29 +01:00
|
|
|
<para>
|
|
|
|
<variablelist>
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-standby-status-update">
|
2011-02-10 20:00:29 +01:00
|
|
|
<term>
|
|
|
|
Standby status update (F)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('r')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a receiver status update.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int64
|
2011-02-10 20:00:29 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The location of the last WAL byte + 1 received and written to disk
|
2012-11-07 17:59:12 +01:00
|
|
|
in the standby.
|
2011-02-10 20:00:29 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int64
|
2011-02-10 20:00:29 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The location of the last WAL byte + 1 flushed to disk in
|
2012-11-07 17:59:12 +01:00
|
|
|
the standby.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int64
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The location of the last WAL byte + 1 applied in the standby.
|
2011-02-10 20:00:29 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int64
|
2011-02-10 20:00:29 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2012-11-07 17:59:12 +01:00
|
|
|
The client's system clock at the time of transmission, as
|
|
|
|
microseconds since midnight on 2000-01-01.
|
2011-02-10 20:00:29 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Byte1
|
2011-02-10 20:00:29 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2012-11-07 17:59:12 +01:00
|
|
|
If 1, the client requests the server to reply to this message
|
|
|
|
immediately. This can be used to ping the server, to test if
|
|
|
|
the connection is still healthy.
|
2011-02-10 20:00:29 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2011-03-17 20:10:55 +01:00
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-hot-standby-feedback-message">
|
2011-03-17 20:10:55 +01:00
|
|
|
<term>
|
2022-03-11 07:16:21 +01:00
|
|
|
Hot standby feedback message (F)
|
2011-03-17 20:10:55 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('h')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2022-03-11 07:16:21 +01:00
|
|
|
Identifies the message as a hot standby feedback message.
|
2011-03-17 20:10:55 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int64
|
2011-03-17 20:10:55 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2012-11-07 17:59:12 +01:00
|
|
|
The client's system clock at the time of transmission, as
|
|
|
|
microseconds since midnight on 2000-01-01.
|
2011-03-17 20:10:55 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int32
|
2011-03-17 20:10:55 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-03-25 15:07:27 +01:00
|
|
|
The standby's current global xmin, excluding the catalog_xmin from any
|
|
|
|
replication slots. If both this value and the following
|
2022-03-11 07:16:21 +01:00
|
|
|
catalog_xmin are 0 this is treated as a notification that hot standby
|
2017-03-25 15:07:27 +01:00
|
|
|
feedback will no longer be sent on this connection. Later non-zero
|
|
|
|
messages may reinitiate the feedback mechanism.
|
2011-03-17 20:10:55 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2012-11-07 17:59:12 +01:00
|
|
|
Int32
|
2011-03-17 20:10:55 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-03-25 15:07:27 +01:00
|
|
|
The epoch of the global xmin xid on the standby.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The lowest catalog_xmin of any replication slots on the standby. Set to 0
|
|
|
|
if no catalog_xmin exists on the standby or if hot standby feedback is being
|
|
|
|
disabled.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The epoch of the catalog_xmin xid on the standby.
|
2011-03-17 20:10:55 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2011-02-10 20:00:29 +01:00
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
2010-06-04 00:17:32 +02:00
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-start-replication-slot-logical">
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><literal>START_REPLICATION</literal> <literal>SLOT</literal> <replaceable class="parameter">slot_name</replaceable> <literal>LOGICAL</literal> <replaceable class="parameter">XXX/XXX</replaceable> [ ( <replaceable>option_name</replaceable> [ <replaceable>option_value</replaceable> ] [, ...] ) ]</term>
|
2014-03-18 18:20:01 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2021-07-30 23:59:19 +02:00
|
|
|
Instructs server to start streaming WAL for logical replication,
|
|
|
|
starting at either WAL location <replaceable
|
|
|
|
class="parameter">XXX/XXX</replaceable> or the slot's
|
|
|
|
<literal>confirmed_flush_lsn</literal> (see <xref
|
|
|
|
linkend="view-pg-replication-slots"/>), whichever is greater. This
|
|
|
|
behavior makes it easier for clients to avoid updating their local LSN
|
|
|
|
status when there is no data to process. However, starting at a
|
|
|
|
different LSN than requested might not catch certain kinds of client
|
|
|
|
errors; so the client may wish to check that
|
|
|
|
<literal>confirmed_flush_lsn</literal> matches its expectations before
|
|
|
|
issuing <literal>START_REPLICATION</literal>.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The server can reply with an error, for example if the
|
2022-01-25 01:40:04 +01:00
|
|
|
slot does not exist. On success, the server responds with a CopyBothResponse
|
2014-03-18 18:20:01 +01:00
|
|
|
message, and then starts to stream WAL to the frontend.
|
|
|
|
</para>
|
2014-05-31 15:58:04 +02:00
|
|
|
|
|
|
|
<para>
|
|
|
|
The messages inside the CopyBothResponse messages are of the same format
|
2020-10-15 01:12:26 +02:00
|
|
|
documented for <literal>START_REPLICATION ... PHYSICAL</literal>, including
|
|
|
|
two CommandComplete messages.
|
2014-05-31 15:58:04 +02:00
|
|
|
</para>
|
|
|
|
|
2014-03-18 18:20:01 +01:00
|
|
|
<para>
|
|
|
|
The output plugin associated with the selected slot is used
|
|
|
|
to process the output for streaming.
|
|
|
|
</para>
|
2014-05-31 15:58:04 +02:00
|
|
|
|
2014-03-18 18:20:01 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><literal>SLOT</literal> <replaceable class="parameter">slot_name</replaceable></term>
|
2014-03-18 18:20:01 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The name of the slot to stream changes from. This parameter is required,
|
|
|
|
and must correspond to an existing logical replication slot created
|
|
|
|
with <literal>CREATE_REPLICATION_SLOT</literal> in
|
|
|
|
<literal>LOGICAL</literal> mode.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><replaceable class="parameter">XXX/XXX</replaceable></term>
|
2014-03-18 18:20:01 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-05-12 19:51:27 +02:00
|
|
|
The WAL location to begin streaming at.
|
2014-03-18 18:20:01 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2014-05-31 15:58:04 +02:00
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><replaceable class="parameter">option_name</replaceable></term>
|
2014-05-31 15:58:04 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The name of an option passed to the slot's logical decoding plugin.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><replaceable class="parameter">option_value</replaceable></term>
|
2014-05-31 15:58:04 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Optional value, in the form of a string constant, associated with the
|
|
|
|
specified option.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2014-03-18 18:20:01 +01:00
|
|
|
</variablelist>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2011-01-10 14:03:55 +01:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-replication-drop-replication-slot">
|
2017-09-01 13:44:14 +02:00
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>DROP_REPLICATION_SLOT</literal> <replaceable class="parameter">slot_name</replaceable> <optional> <literal>WAIT</literal> </optional>
|
2014-08-18 04:18:53 +02:00
|
|
|
<indexterm><primary>DROP_REPLICATION_SLOT</primary></indexterm>
|
|
|
|
</term>
|
2014-02-01 04:45:17 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-09-01 13:44:14 +02:00
|
|
|
Drops a replication slot, freeing any reserved server-side resources.
|
2017-03-28 16:05:21 +02:00
|
|
|
If the slot is a logical slot that was created in a database other than
|
|
|
|
the database the walsender is connected to, this command fails.
|
2014-02-01 04:45:17 +01:00
|
|
|
</para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><replaceable class="parameter">slot_name</replaceable></term>
|
2014-02-01 04:45:17 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The name of the slot to drop.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2017-09-01 13:44:14 +02:00
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term><literal>WAIT</literal></term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
This option causes the command to wait if the slot is active until
|
|
|
|
it becomes inactive, instead of the default behavior of raising an
|
|
|
|
error.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2014-02-01 04:45:17 +01:00
|
|
|
</variablelist>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2020-03-03 04:03:43 +01:00
|
|
|
<varlistentry id="protocol-replication-base-backup" xreflabel="BASE_BACKUP">
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
<term><literal>BASE_BACKUP</literal> [ ( <replaceable class="parameter">option</replaceable> [, ...] ) ]
|
2014-08-18 04:18:53 +02:00
|
|
|
<indexterm><primary>BASE_BACKUP</primary></indexterm>
|
|
|
|
</term>
|
2011-01-10 14:03:55 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Instructs the server to start streaming a base backup.
|
2011-01-14 16:30:33 +01:00
|
|
|
The system will automatically be put in backup mode before the backup
|
|
|
|
is started, and taken out of it when the backup is complete. The
|
|
|
|
following options are accepted:
|
2011-01-10 14:03:55 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
2011-01-14 16:30:33 +01:00
|
|
|
<term><literal>LABEL</literal> <replaceable>'label'</replaceable></term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Sets the label of the backup. If none is specified, a backup label
|
|
|
|
of <literal>base backup</literal> will be used. The quoting rules
|
|
|
|
for the label are the same as a standard SQL string with
|
2017-11-23 15:39:47 +01:00
|
|
|
<xref linkend="guc-standard-conforming-strings"/> turned on.
|
2011-01-14 16:30:33 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
Modify pg_basebackup to use a new COPY subprotocol for base backups.
In the new approach, all files across all tablespaces are sent in a
single COPY OUT operation. The CopyData messages are no longer raw
archive content; rather, each message is prefixed with a type byte
that describes its purpose, e.g. 'n' signifies the start of a new
archive and 'd' signifies archive or manifest data. This protocol
is significantly more extensible than the old approach, since we can
later create more message types, though not without concern for
backward compatibility.
The new protocol sends a few things to the client that the old one
did not. First, it sends the name of each archive explicitly, instead
of letting the client compute it. This is intended to make it easier
to write future patches that might send archives in a format other
that tar (e.g. cpio, pax, tar.gz). Second, it sends explicit progress
messages rather than allowing the client to assume that progress is
defined by the number of bytes received. This will help with future
features where the server compresses the data, or sends it someplace
directly rather than transmitting it to the client.
The old protocol is still supported for compatibility with previous
releases. The new protocol is selected by means of a new
TARGET option to the BASE_BACKUP command. Currently, the
only supported target is 'client'. Support for additional
targets will be added in a later commit.
Patch by me. The patch set of which this is a part has had review
and/or testing from Jeevan Ladhe, Tushar Ahuja, Suraj Kharage,
Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+TgmoaYZbz0=Yk797aOJwkGJC-LK3iXn+wzzMx7KdwNpZhS5g@mail.gmail.com
2022-01-18 19:47:26 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term><literal>TARGET</literal> <replaceable>'target'</replaceable></term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2022-02-08 17:17:21 +01:00
|
|
|
Tells the server where to send the backup. If the target is
|
|
|
|
<literal>client</literal>, which is the default, the backup data is
|
Support base backup targets.
pg_basebackup now has a --target=TARGET[:DETAIL] option. If specfied,
it is sent to the server as the value of the TARGET option to the
BASE_BACKUP command. If DETAIL is included, it is sent as the value of
the new TARGET_DETAIL option to the BASE_BACKUP command. If the
target is anything other than 'client', pg_basebackup assumes that it
will now be the server's job to write the backup in a location somehow
defined by the target, and that it therefore needs to write nothing
locally. However, the server will still send messages to the client
for progress reporting purposes.
On the server side, we now support two additional types of backup
targets. There is a 'blackhole' target, which just throws away the
backup data without doing anything at all with it. Naturally, this
should only be used for testing and debugging purposes, since you will
not actually have a backup when it finishes running. More usefully,
there is also a 'server' target, so you can now use something like
'pg_basebackup -Xnone -t server:/SOME/PATH' to write a backup to some
location on the server. We can extend this to more types of targets
in the future, and might even want to create an extensibility
mechanism for adding new target types.
Since WAL fetching is handled with separate client-side logic, it's
not part of this mechanism; thus, backups with non-default targets
must use -Xnone or -Xfetch.
Patch by me, with a bug fix by Jeevan Ladhe. The patch set of which
this is a part has also had review and/or testing from Tushar Ahuja,
Suraj Kharage, Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+TgmoaYZbz0=Yk797aOJwkGJC-LK3iXn+wzzMx7KdwNpZhS5g@mail.gmail.com
2021-11-16 21:20:50 +01:00
|
|
|
sent to the client. If it is <literal>server</literal>, the backup
|
|
|
|
data is written to the server at the pathname specified by the
|
|
|
|
<literal>TARGET_DETAIL</literal> option. If it is
|
|
|
|
<literal>blackhole</literal>, the backup data is not sent
|
|
|
|
anywhere; it is simply discarded.
|
|
|
|
</para>
|
2022-01-28 18:26:33 +01:00
|
|
|
|
|
|
|
<para>
|
|
|
|
The <literal>server</literal> target requires superuser privilege or
|
|
|
|
being granted the <literal>pg_write_server_files</literal> role.
|
|
|
|
</para>
|
Support base backup targets.
pg_basebackup now has a --target=TARGET[:DETAIL] option. If specfied,
it is sent to the server as the value of the TARGET option to the
BASE_BACKUP command. If DETAIL is included, it is sent as the value of
the new TARGET_DETAIL option to the BASE_BACKUP command. If the
target is anything other than 'client', pg_basebackup assumes that it
will now be the server's job to write the backup in a location somehow
defined by the target, and that it therefore needs to write nothing
locally. However, the server will still send messages to the client
for progress reporting purposes.
On the server side, we now support two additional types of backup
targets. There is a 'blackhole' target, which just throws away the
backup data without doing anything at all with it. Naturally, this
should only be used for testing and debugging purposes, since you will
not actually have a backup when it finishes running. More usefully,
there is also a 'server' target, so you can now use something like
'pg_basebackup -Xnone -t server:/SOME/PATH' to write a backup to some
location on the server. We can extend this to more types of targets
in the future, and might even want to create an extensibility
mechanism for adding new target types.
Since WAL fetching is handled with separate client-side logic, it's
not part of this mechanism; thus, backups with non-default targets
must use -Xnone or -Xfetch.
Patch by me, with a bug fix by Jeevan Ladhe. The patch set of which
this is a part has also had review and/or testing from Tushar Ahuja,
Suraj Kharage, Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+TgmoaYZbz0=Yk797aOJwkGJC-LK3iXn+wzzMx7KdwNpZhS5g@mail.gmail.com
2021-11-16 21:20:50 +01:00
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term><literal>TARGET_DETAIL</literal> <replaceable>'detail'</replaceable></term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Provides additional information about the backup target.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Currently, this option can only be used when the backup target is
|
|
|
|
<literal>server</literal>. It specifies the server directory
|
|
|
|
to which the backup should be written.
|
Modify pg_basebackup to use a new COPY subprotocol for base backups.
In the new approach, all files across all tablespaces are sent in a
single COPY OUT operation. The CopyData messages are no longer raw
archive content; rather, each message is prefixed with a type byte
that describes its purpose, e.g. 'n' signifies the start of a new
archive and 'd' signifies archive or manifest data. This protocol
is significantly more extensible than the old approach, since we can
later create more message types, though not without concern for
backward compatibility.
The new protocol sends a few things to the client that the old one
did not. First, it sends the name of each archive explicitly, instead
of letting the client compute it. This is intended to make it easier
to write future patches that might send archives in a format other
that tar (e.g. cpio, pax, tar.gz). Second, it sends explicit progress
messages rather than allowing the client to assume that progress is
defined by the number of bytes received. This will help with future
features where the server compresses the data, or sends it someplace
directly rather than transmitting it to the client.
The old protocol is still supported for compatibility with previous
releases. The new protocol is selected by means of a new
TARGET option to the BASE_BACKUP command. Currently, the
only supported target is 'client'. Support for additional
targets will be added in a later commit.
Patch by me. The patch set of which this is a part has had review
and/or testing from Jeevan Ladhe, Tushar Ahuja, Suraj Kharage,
Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+TgmoaYZbz0=Yk797aOJwkGJC-LK3iXn+wzzMx7KdwNpZhS5g@mail.gmail.com
2022-01-18 19:47:26 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2011-01-14 16:30:33 +01:00
|
|
|
<varlistentry>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
<term><literal>PROGRESS [ <replaceable class="parameter">boolean</replaceable> ]</literal></term>
|
2011-01-10 14:03:55 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
If set to true, request information required to generate a progress
|
|
|
|
report. This will send back an approximate size in the header of each
|
|
|
|
tablespace, which can be used to calculate how far along the stream
|
|
|
|
is done. This is calculated by enumerating all the file sizes once
|
|
|
|
before the transfer is even started, and might as such have a
|
|
|
|
negative impact on the performance. In particular, it might take
|
|
|
|
longer before the first data
|
2011-01-10 14:03:55 +01:00
|
|
|
is streamed. Since the database files can change during the backup,
|
2015-10-22 04:31:56 +02:00
|
|
|
the size is only approximate and might both grow and shrink between
|
2011-01-10 14:03:55 +01:00
|
|
|
the time of approximation and the sending of the actual files.
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
The default is false.
|
2011-01-10 14:03:55 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2011-01-23 12:21:23 +01:00
|
|
|
|
|
|
|
<varlistentry>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
<term><literal>CHECKPOINT { 'fast' | 'spread' }</literal></term>
|
2011-01-23 12:21:23 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
Sets the type of checkpoint to be performed at the beginning of the
|
|
|
|
base backup. The default is <literal>spread</literal>.
|
2011-01-23 12:21:23 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2011-01-30 21:30:09 +01:00
|
|
|
|
|
|
|
<varlistentry>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
<term><literal>WAL [ <replaceable class="parameter">boolean</replaceable> ]</literal></term>
|
2011-01-30 21:30:09 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
If set to true, include the necessary WAL segments in the backup.
|
|
|
|
This will include all the files between start and stop backup in the
|
2016-10-20 17:24:37 +02:00
|
|
|
<filename>pg_wal</filename> directory of the base directory tar
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
file. The default is false.
|
2011-01-30 21:30:09 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
2011-02-09 10:59:53 +01:00
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
<term><literal>WAIT [ <replaceable class="parameter">boolean</replaceable> ]</literal></term>
|
2011-02-09 10:59:53 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
If set to true, the backup will wait until the last required WAL
|
2011-02-09 10:59:53 +01:00
|
|
|
segment has been archived, or emit a warning if log archiving is
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
not enabled. If false, the backup will neither wait nor warn,
|
|
|
|
leaving the client responsible for ensuring the required log is
|
|
|
|
available. The default is true.
|
2011-02-09 10:59:53 +01:00
|
|
|
</para>
|
2014-02-27 22:55:57 +01:00
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
Server-side gzip compression.
pg_basebackup's --compression option now lets you write either
"client-gzip" or "server-gzip" instead of just "gzip" to specify
where the compression should be performed. If you write simply
"gzip" it's taken to mean "client-gzip" unless you also use
--target, in which case it is interpreted to mean "server-gzip",
because that's the only thing that makes any sense in that case.
To make this work, the BASE_BACKUP command now takes new
COMPRESSION and COMPRESSION_LEVEL options.
At present, pg_basebackup cannot decompress .gz files, so
server-side compression will cause a failure if (1) -Ft is not
used or (2) -R is used or (3) -D- is used without --no-manifest.
Along the way, I removed the information message added by commit
5c649fe153367cdab278738ee4aebbfd158e0546 which occurred if you
specified no compression level and told you that the default level
had been used instead. That seemed like more output than most
people would want.
Also along the way, this adds a check to the server for
unrecognized base backup options. This repairs a bug introduced
by commit 0ba281cb4bf9f5f65529dfa4c8282abb734dd454.
This commit also adds some new test cases for pg_verifybackup.
They take a server-side backup with and without compression, and
then extract the backup if we have the OS facilities available
to do so, and then run pg_verifybackup on the extracted
directory. That is a good test of the functionality added by
this commit and also improves test coverage for the backup target
patch (commit 3500ccc39b0dadd1068a03938e4b8ff562587ccc) and for
pg_verifybackup itself.
Patch by me, with a bug fix by Jeevan Ladhe. The patch set of which
this is a part has also had review and/or testing from Tushar Ahuja,
Suraj Kharage, Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+Tgmoa-ST7fMLsVJduOB7Eub=2WjfpHS+QxHVEpUoinf4bOSg@mail.gmail.com
2022-01-24 21:13:18 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term><literal>COMPRESSION</literal> <replaceable>'method'</replaceable></term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Instructs the server to compress the backup using the specified
|
2022-03-07 21:08:45 +01:00
|
|
|
method. Currently, the supported methods are <literal>gzip</literal>,
|
|
|
|
<literal>lz4</literal>, and <literal>zstd</literal>.
|
Server-side gzip compression.
pg_basebackup's --compression option now lets you write either
"client-gzip" or "server-gzip" instead of just "gzip" to specify
where the compression should be performed. If you write simply
"gzip" it's taken to mean "client-gzip" unless you also use
--target, in which case it is interpreted to mean "server-gzip",
because that's the only thing that makes any sense in that case.
To make this work, the BASE_BACKUP command now takes new
COMPRESSION and COMPRESSION_LEVEL options.
At present, pg_basebackup cannot decompress .gz files, so
server-side compression will cause a failure if (1) -Ft is not
used or (2) -R is used or (3) -D- is used without --no-manifest.
Along the way, I removed the information message added by commit
5c649fe153367cdab278738ee4aebbfd158e0546 which occurred if you
specified no compression level and told you that the default level
had been used instead. That seemed like more output than most
people would want.
Also along the way, this adds a check to the server for
unrecognized base backup options. This repairs a bug introduced
by commit 0ba281cb4bf9f5f65529dfa4c8282abb734dd454.
This commit also adds some new test cases for pg_verifybackup.
They take a server-side backup with and without compression, and
then extract the backup if we have the OS facilities available
to do so, and then run pg_verifybackup on the extracted
directory. That is a good test of the functionality added by
this commit and also improves test coverage for the backup target
patch (commit 3500ccc39b0dadd1068a03938e4b8ff562587ccc) and for
pg_verifybackup itself.
Patch by me, with a bug fix by Jeevan Ladhe. The patch set of which
this is a part has also had review and/or testing from Tushar Ahuja,
Suraj Kharage, Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+Tgmoa-ST7fMLsVJduOB7Eub=2WjfpHS+QxHVEpUoinf4bOSg@mail.gmail.com
2022-01-24 21:13:18 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2022-03-23 14:19:14 +01:00
|
|
|
<term><literal>COMPRESSION_DETAIL</literal> <replaceable>detail</replaceable></term>
|
Server-side gzip compression.
pg_basebackup's --compression option now lets you write either
"client-gzip" or "server-gzip" instead of just "gzip" to specify
where the compression should be performed. If you write simply
"gzip" it's taken to mean "client-gzip" unless you also use
--target, in which case it is interpreted to mean "server-gzip",
because that's the only thing that makes any sense in that case.
To make this work, the BASE_BACKUP command now takes new
COMPRESSION and COMPRESSION_LEVEL options.
At present, pg_basebackup cannot decompress .gz files, so
server-side compression will cause a failure if (1) -Ft is not
used or (2) -R is used or (3) -D- is used without --no-manifest.
Along the way, I removed the information message added by commit
5c649fe153367cdab278738ee4aebbfd158e0546 which occurred if you
specified no compression level and told you that the default level
had been used instead. That seemed like more output than most
people would want.
Also along the way, this adds a check to the server for
unrecognized base backup options. This repairs a bug introduced
by commit 0ba281cb4bf9f5f65529dfa4c8282abb734dd454.
This commit also adds some new test cases for pg_verifybackup.
They take a server-side backup with and without compression, and
then extract the backup if we have the OS facilities available
to do so, and then run pg_verifybackup on the extracted
directory. That is a good test of the functionality added by
this commit and also improves test coverage for the backup target
patch (commit 3500ccc39b0dadd1068a03938e4b8ff562587ccc) and for
pg_verifybackup itself.
Patch by me, with a bug fix by Jeevan Ladhe. The patch set of which
this is a part has also had review and/or testing from Tushar Ahuja,
Suraj Kharage, Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+Tgmoa-ST7fMLsVJduOB7Eub=2WjfpHS+QxHVEpUoinf4bOSg@mail.gmail.com
2022-01-24 21:13:18 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2022-03-23 14:19:14 +01:00
|
|
|
Specifies details for the chosen compression method. This should only
|
|
|
|
be used in conjunction with the <literal>COMPRESSION</literal>
|
|
|
|
option. If the value is an integer, it specifies the compression
|
|
|
|
level. Otherwise, it should be a comma-separated list of items,
|
|
|
|
each of the form <literal>keyword</literal> or
|
2022-03-30 15:35:14 +02:00
|
|
|
<literal>keyword=value</literal>. Currently, the supported keywords
|
|
|
|
are <literal>level</literal> and <literal>workers</literal>.
|
2022-03-23 14:19:14 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2022-03-30 15:35:14 +02:00
|
|
|
The <literal>level</literal> keyword sets the compression level.
|
2022-03-23 14:19:14 +01:00
|
|
|
For <literal>gzip</literal> the compression level should be an
|
|
|
|
integer between 1 and 9, for <literal>lz4</literal> an integer
|
|
|
|
between 1 and 12, and for <literal>zstd</literal> an integer
|
|
|
|
between 1 and 22.
|
Server-side gzip compression.
pg_basebackup's --compression option now lets you write either
"client-gzip" or "server-gzip" instead of just "gzip" to specify
where the compression should be performed. If you write simply
"gzip" it's taken to mean "client-gzip" unless you also use
--target, in which case it is interpreted to mean "server-gzip",
because that's the only thing that makes any sense in that case.
To make this work, the BASE_BACKUP command now takes new
COMPRESSION and COMPRESSION_LEVEL options.
At present, pg_basebackup cannot decompress .gz files, so
server-side compression will cause a failure if (1) -Ft is not
used or (2) -R is used or (3) -D- is used without --no-manifest.
Along the way, I removed the information message added by commit
5c649fe153367cdab278738ee4aebbfd158e0546 which occurred if you
specified no compression level and told you that the default level
had been used instead. That seemed like more output than most
people would want.
Also along the way, this adds a check to the server for
unrecognized base backup options. This repairs a bug introduced
by commit 0ba281cb4bf9f5f65529dfa4c8282abb734dd454.
This commit also adds some new test cases for pg_verifybackup.
They take a server-side backup with and without compression, and
then extract the backup if we have the OS facilities available
to do so, and then run pg_verifybackup on the extracted
directory. That is a good test of the functionality added by
this commit and also improves test coverage for the backup target
patch (commit 3500ccc39b0dadd1068a03938e4b8ff562587ccc) and for
pg_verifybackup itself.
Patch by me, with a bug fix by Jeevan Ladhe. The patch set of which
this is a part has also had review and/or testing from Tushar Ahuja,
Suraj Kharage, Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+Tgmoa-ST7fMLsVJduOB7Eub=2WjfpHS+QxHVEpUoinf4bOSg@mail.gmail.com
2022-01-24 21:13:18 +01:00
|
|
|
</para>
|
2022-03-30 15:35:14 +02:00
|
|
|
|
|
|
|
<para>
|
|
|
|
The <literal>workers</literal> keyword sets the number of threads
|
|
|
|
that should be used for parallel compression. Parallel compression
|
|
|
|
is supported only for <literal>zstd</literal>.
|
|
|
|
</para>
|
Server-side gzip compression.
pg_basebackup's --compression option now lets you write either
"client-gzip" or "server-gzip" instead of just "gzip" to specify
where the compression should be performed. If you write simply
"gzip" it's taken to mean "client-gzip" unless you also use
--target, in which case it is interpreted to mean "server-gzip",
because that's the only thing that makes any sense in that case.
To make this work, the BASE_BACKUP command now takes new
COMPRESSION and COMPRESSION_LEVEL options.
At present, pg_basebackup cannot decompress .gz files, so
server-side compression will cause a failure if (1) -Ft is not
used or (2) -R is used or (3) -D- is used without --no-manifest.
Along the way, I removed the information message added by commit
5c649fe153367cdab278738ee4aebbfd158e0546 which occurred if you
specified no compression level and told you that the default level
had been used instead. That seemed like more output than most
people would want.
Also along the way, this adds a check to the server for
unrecognized base backup options. This repairs a bug introduced
by commit 0ba281cb4bf9f5f65529dfa4c8282abb734dd454.
This commit also adds some new test cases for pg_verifybackup.
They take a server-side backup with and without compression, and
then extract the backup if we have the OS facilities available
to do so, and then run pg_verifybackup on the extracted
directory. That is a good test of the functionality added by
this commit and also improves test coverage for the backup target
patch (commit 3500ccc39b0dadd1068a03938e4b8ff562587ccc) and for
pg_verifybackup itself.
Patch by me, with a bug fix by Jeevan Ladhe. The patch set of which
this is a part has also had review and/or testing from Tushar Ahuja,
Suraj Kharage, Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+Tgmoa-ST7fMLsVJduOB7Eub=2WjfpHS+QxHVEpUoinf4bOSg@mail.gmail.com
2022-01-24 21:13:18 +01:00
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2014-02-27 22:55:57 +01:00
|
|
|
<varlistentry>
|
2017-10-09 03:44:17 +02:00
|
|
|
<term><literal>MAX_RATE</literal> <replaceable>rate</replaceable></term>
|
2014-02-27 22:55:57 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Limit (throttle) the maximum amount of data transferred from server
|
|
|
|
to client per unit of time. The expected unit is kilobytes per second.
|
|
|
|
If this option is specified, the value must either be equal to zero
|
2016-08-05 20:35:09 +02:00
|
|
|
or it must fall within the range from 32 kB through 1 GB (inclusive).
|
2014-02-27 22:55:57 +01:00
|
|
|
If zero is passed or the option is not specified, no restriction is
|
|
|
|
imposed on the transfer.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
2011-01-30 21:30:09 +01:00
|
|
|
</varlistentry>
|
2015-05-12 15:29:10 +02:00
|
|
|
|
|
|
|
<varlistentry>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
<term><literal>TABLESPACE_MAP [ <replaceable class="parameter">boolean</replaceable> ]</literal></term>
|
2015-05-12 15:29:10 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
If true, include information about symbolic links present in the
|
|
|
|
directory <filename>pg_tblspc</filename> in a file named
|
2015-05-12 15:29:10 +02:00
|
|
|
<filename>tablespace_map</filename>. The tablespace map file includes
|
|
|
|
each symbolic link name as it exists in the directory
|
2017-10-09 03:44:17 +02:00
|
|
|
<filename>pg_tblspc/</filename> and the full path of that symbolic link.
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
The default is false.
|
2015-05-12 15:29:10 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2018-04-03 13:47:16 +02:00
|
|
|
|
|
|
|
<varlistentry>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
<term><literal>VERIFY_CHECKSUMS [ <replaceable class="parameter">boolean</replaceable> ]</literal></term>
|
2018-04-03 13:47:16 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
If true, checksums are verified during a base backup if they are
|
|
|
|
enabled. If false, this is skipped. The default is true.
|
2018-04-03 13:47:16 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
Generate backup manifests for base backups, and validate them.
A manifest is a JSON document which includes (1) the file name, size,
last modification time, and an optional checksum for each file backed
up, (2) timelines and LSNs for whatever WAL will need to be replayed
to make the backup consistent, and (3) a checksum for the manifest
itself. By default, we use CRC-32C when checksumming data files,
because we are trying to detect corruption and user error, not foil an
adversary. However, pg_basebackup and the server-side BASE_BACKUP
command now have options to select a different algorithm, so users
wanting a cryptographic hash function can select SHA-224, SHA-256,
SHA-384, or SHA-512. Users not wanting file checksums at all can
disable them, or disable generating of the backup manifest altogether.
Using a cryptographic hash function in place of CRC-32C consumes
significantly more CPU cycles, which may slow down backups in some
cases.
A new tool called pg_validatebackup can validate a backup against the
manifest. If no checksums are present, it can still check that the
right files exist and that they have the expected sizes. If checksums
are present, it can also verify that each file has the expected
checksum. Additionally, it calls pg_waldump to verify that the
expected WAL files are present and parseable. Only plain format
backups can be validated directly, but tar format backups can be
validated after extracting them.
Robert Haas, with help, ideas, review, and testing from David Steele,
Stephen Frost, Andrew Dunstan, Rushabh Lathia, Suraj Kharage, Tushar
Ahuja, Rajkumar Raghuwanshi, Mark Dilger, Davinder Singh, Jeevan
Chalke, Amit Kapila, Andres Freund, and Noah Misch.
Discussion: http://postgr.es/m/CA+TgmoZV8dw1H2bzZ9xkKwdrk8+XYa+DC9H=F7heO2zna5T6qg@mail.gmail.com
2020-04-03 20:59:47 +02:00
|
|
|
|
|
|
|
<varlistentry>
|
2020-04-15 04:15:12 +02:00
|
|
|
<term><literal>MANIFEST</literal> <replaceable>manifest_option</replaceable></term>
|
Generate backup manifests for base backups, and validate them.
A manifest is a JSON document which includes (1) the file name, size,
last modification time, and an optional checksum for each file backed
up, (2) timelines and LSNs for whatever WAL will need to be replayed
to make the backup consistent, and (3) a checksum for the manifest
itself. By default, we use CRC-32C when checksumming data files,
because we are trying to detect corruption and user error, not foil an
adversary. However, pg_basebackup and the server-side BASE_BACKUP
command now have options to select a different algorithm, so users
wanting a cryptographic hash function can select SHA-224, SHA-256,
SHA-384, or SHA-512. Users not wanting file checksums at all can
disable them, or disable generating of the backup manifest altogether.
Using a cryptographic hash function in place of CRC-32C consumes
significantly more CPU cycles, which may slow down backups in some
cases.
A new tool called pg_validatebackup can validate a backup against the
manifest. If no checksums are present, it can still check that the
right files exist and that they have the expected sizes. If checksums
are present, it can also verify that each file has the expected
checksum. Additionally, it calls pg_waldump to verify that the
expected WAL files are present and parseable. Only plain format
backups can be validated directly, but tar format backups can be
validated after extracting them.
Robert Haas, with help, ideas, review, and testing from David Steele,
Stephen Frost, Andrew Dunstan, Rushabh Lathia, Suraj Kharage, Tushar
Ahuja, Rajkumar Raghuwanshi, Mark Dilger, Davinder Singh, Jeevan
Chalke, Amit Kapila, Andres Freund, and Noah Misch.
Discussion: http://postgr.es/m/CA+TgmoZV8dw1H2bzZ9xkKwdrk8+XYa+DC9H=F7heO2zna5T6qg@mail.gmail.com
2020-04-03 20:59:47 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
When this option is specified with a value of <literal>yes</literal>
|
2020-04-15 04:15:12 +02:00
|
|
|
or <literal>force-encode</literal>, a backup manifest is created
|
Generate backup manifests for base backups, and validate them.
A manifest is a JSON document which includes (1) the file name, size,
last modification time, and an optional checksum for each file backed
up, (2) timelines and LSNs for whatever WAL will need to be replayed
to make the backup consistent, and (3) a checksum for the manifest
itself. By default, we use CRC-32C when checksumming data files,
because we are trying to detect corruption and user error, not foil an
adversary. However, pg_basebackup and the server-side BASE_BACKUP
command now have options to select a different algorithm, so users
wanting a cryptographic hash function can select SHA-224, SHA-256,
SHA-384, or SHA-512. Users not wanting file checksums at all can
disable them, or disable generating of the backup manifest altogether.
Using a cryptographic hash function in place of CRC-32C consumes
significantly more CPU cycles, which may slow down backups in some
cases.
A new tool called pg_validatebackup can validate a backup against the
manifest. If no checksums are present, it can still check that the
right files exist and that they have the expected sizes. If checksums
are present, it can also verify that each file has the expected
checksum. Additionally, it calls pg_waldump to verify that the
expected WAL files are present and parseable. Only plain format
backups can be validated directly, but tar format backups can be
validated after extracting them.
Robert Haas, with help, ideas, review, and testing from David Steele,
Stephen Frost, Andrew Dunstan, Rushabh Lathia, Suraj Kharage, Tushar
Ahuja, Rajkumar Raghuwanshi, Mark Dilger, Davinder Singh, Jeevan
Chalke, Amit Kapila, Andres Freund, and Noah Misch.
Discussion: http://postgr.es/m/CA+TgmoZV8dw1H2bzZ9xkKwdrk8+XYa+DC9H=F7heO2zna5T6qg@mail.gmail.com
2020-04-03 20:59:47 +02:00
|
|
|
and sent along with the backup. The manifest is a list of every
|
|
|
|
file present in the backup with the exception of any WAL files that
|
|
|
|
may be included. It also stores the size, last modification time, and
|
2020-09-21 18:43:42 +02:00
|
|
|
optionally a checksum for each file.
|
2020-04-15 04:15:12 +02:00
|
|
|
A value of <literal>force-encode</literal> forces all filenames
|
Generate backup manifests for base backups, and validate them.
A manifest is a JSON document which includes (1) the file name, size,
last modification time, and an optional checksum for each file backed
up, (2) timelines and LSNs for whatever WAL will need to be replayed
to make the backup consistent, and (3) a checksum for the manifest
itself. By default, we use CRC-32C when checksumming data files,
because we are trying to detect corruption and user error, not foil an
adversary. However, pg_basebackup and the server-side BASE_BACKUP
command now have options to select a different algorithm, so users
wanting a cryptographic hash function can select SHA-224, SHA-256,
SHA-384, or SHA-512. Users not wanting file checksums at all can
disable them, or disable generating of the backup manifest altogether.
Using a cryptographic hash function in place of CRC-32C consumes
significantly more CPU cycles, which may slow down backups in some
cases.
A new tool called pg_validatebackup can validate a backup against the
manifest. If no checksums are present, it can still check that the
right files exist and that they have the expected sizes. If checksums
are present, it can also verify that each file has the expected
checksum. Additionally, it calls pg_waldump to verify that the
expected WAL files are present and parseable. Only plain format
backups can be validated directly, but tar format backups can be
validated after extracting them.
Robert Haas, with help, ideas, review, and testing from David Steele,
Stephen Frost, Andrew Dunstan, Rushabh Lathia, Suraj Kharage, Tushar
Ahuja, Rajkumar Raghuwanshi, Mark Dilger, Davinder Singh, Jeevan
Chalke, Amit Kapila, Andres Freund, and Noah Misch.
Discussion: http://postgr.es/m/CA+TgmoZV8dw1H2bzZ9xkKwdrk8+XYa+DC9H=F7heO2zna5T6qg@mail.gmail.com
2020-04-03 20:59:47 +02:00
|
|
|
to be hex-encoded; otherwise, this type of encoding is performed only
|
|
|
|
for files whose names are non-UTF8 octet sequences.
|
2020-04-15 04:15:12 +02:00
|
|
|
<literal>force-encode</literal> is intended primarily for testing
|
Generate backup manifests for base backups, and validate them.
A manifest is a JSON document which includes (1) the file name, size,
last modification time, and an optional checksum for each file backed
up, (2) timelines and LSNs for whatever WAL will need to be replayed
to make the backup consistent, and (3) a checksum for the manifest
itself. By default, we use CRC-32C when checksumming data files,
because we are trying to detect corruption and user error, not foil an
adversary. However, pg_basebackup and the server-side BASE_BACKUP
command now have options to select a different algorithm, so users
wanting a cryptographic hash function can select SHA-224, SHA-256,
SHA-384, or SHA-512. Users not wanting file checksums at all can
disable them, or disable generating of the backup manifest altogether.
Using a cryptographic hash function in place of CRC-32C consumes
significantly more CPU cycles, which may slow down backups in some
cases.
A new tool called pg_validatebackup can validate a backup against the
manifest. If no checksums are present, it can still check that the
right files exist and that they have the expected sizes. If checksums
are present, it can also verify that each file has the expected
checksum. Additionally, it calls pg_waldump to verify that the
expected WAL files are present and parseable. Only plain format
backups can be validated directly, but tar format backups can be
validated after extracting them.
Robert Haas, with help, ideas, review, and testing from David Steele,
Stephen Frost, Andrew Dunstan, Rushabh Lathia, Suraj Kharage, Tushar
Ahuja, Rajkumar Raghuwanshi, Mark Dilger, Davinder Singh, Jeevan
Chalke, Amit Kapila, Andres Freund, and Noah Misch.
Discussion: http://postgr.es/m/CA+TgmoZV8dw1H2bzZ9xkKwdrk8+XYa+DC9H=F7heO2zna5T6qg@mail.gmail.com
2020-04-03 20:59:47 +02:00
|
|
|
purposes, to be sure that clients which read the backup manifest
|
|
|
|
can handle this case. For compatibility with previous releases,
|
|
|
|
the default is <literal>MANIFEST 'no'</literal>.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2020-04-15 04:15:12 +02:00
|
|
|
<term><literal>MANIFEST_CHECKSUMS</literal> <replaceable>checksum_algorithm</replaceable></term>
|
Generate backup manifests for base backups, and validate them.
A manifest is a JSON document which includes (1) the file name, size,
last modification time, and an optional checksum for each file backed
up, (2) timelines and LSNs for whatever WAL will need to be replayed
to make the backup consistent, and (3) a checksum for the manifest
itself. By default, we use CRC-32C when checksumming data files,
because we are trying to detect corruption and user error, not foil an
adversary. However, pg_basebackup and the server-side BASE_BACKUP
command now have options to select a different algorithm, so users
wanting a cryptographic hash function can select SHA-224, SHA-256,
SHA-384, or SHA-512. Users not wanting file checksums at all can
disable them, or disable generating of the backup manifest altogether.
Using a cryptographic hash function in place of CRC-32C consumes
significantly more CPU cycles, which may slow down backups in some
cases.
A new tool called pg_validatebackup can validate a backup against the
manifest. If no checksums are present, it can still check that the
right files exist and that they have the expected sizes. If checksums
are present, it can also verify that each file has the expected
checksum. Additionally, it calls pg_waldump to verify that the
expected WAL files are present and parseable. Only plain format
backups can be validated directly, but tar format backups can be
validated after extracting them.
Robert Haas, with help, ideas, review, and testing from David Steele,
Stephen Frost, Andrew Dunstan, Rushabh Lathia, Suraj Kharage, Tushar
Ahuja, Rajkumar Raghuwanshi, Mark Dilger, Davinder Singh, Jeevan
Chalke, Amit Kapila, Andres Freund, and Noah Misch.
Discussion: http://postgr.es/m/CA+TgmoZV8dw1H2bzZ9xkKwdrk8+XYa+DC9H=F7heO2zna5T6qg@mail.gmail.com
2020-04-03 20:59:47 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2020-09-21 18:43:42 +02:00
|
|
|
Specifies the checksum algorithm that should be applied to each file included
|
Generate backup manifests for base backups, and validate them.
A manifest is a JSON document which includes (1) the file name, size,
last modification time, and an optional checksum for each file backed
up, (2) timelines and LSNs for whatever WAL will need to be replayed
to make the backup consistent, and (3) a checksum for the manifest
itself. By default, we use CRC-32C when checksumming data files,
because we are trying to detect corruption and user error, not foil an
adversary. However, pg_basebackup and the server-side BASE_BACKUP
command now have options to select a different algorithm, so users
wanting a cryptographic hash function can select SHA-224, SHA-256,
SHA-384, or SHA-512. Users not wanting file checksums at all can
disable them, or disable generating of the backup manifest altogether.
Using a cryptographic hash function in place of CRC-32C consumes
significantly more CPU cycles, which may slow down backups in some
cases.
A new tool called pg_validatebackup can validate a backup against the
manifest. If no checksums are present, it can still check that the
right files exist and that they have the expected sizes. If checksums
are present, it can also verify that each file has the expected
checksum. Additionally, it calls pg_waldump to verify that the
expected WAL files are present and parseable. Only plain format
backups can be validated directly, but tar format backups can be
validated after extracting them.
Robert Haas, with help, ideas, review, and testing from David Steele,
Stephen Frost, Andrew Dunstan, Rushabh Lathia, Suraj Kharage, Tushar
Ahuja, Rajkumar Raghuwanshi, Mark Dilger, Davinder Singh, Jeevan
Chalke, Amit Kapila, Andres Freund, and Noah Misch.
Discussion: http://postgr.es/m/CA+TgmoZV8dw1H2bzZ9xkKwdrk8+XYa+DC9H=F7heO2zna5T6qg@mail.gmail.com
2020-04-03 20:59:47 +02:00
|
|
|
in the backup manifest. Currently, the available
|
|
|
|
algorithms are <literal>NONE</literal>, <literal>CRC32C</literal>,
|
|
|
|
<literal>SHA224</literal>, <literal>SHA256</literal>,
|
|
|
|
<literal>SHA384</literal>, and <literal>SHA512</literal>.
|
|
|
|
The default is <literal>CRC32C</literal>.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2011-01-10 14:03:55 +01:00
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
<para>
|
2011-02-03 13:46:23 +01:00
|
|
|
When the backup is started, the server will first send two
|
2021-01-16 23:40:12 +01:00
|
|
|
ordinary result sets, followed by one or more CopyOutResponse
|
2011-02-03 13:46:23 +01:00
|
|
|
results.
|
|
|
|
</para>
|
|
|
|
<para>
|
|
|
|
The first ordinary result set contains the starting position of the
|
Make pg_receivexlog and pg_basebackup -X stream work across timeline switches.
This mirrors the changes done earlier to the server in standby mode. When
receivelog reaches the end of a timeline, as reported by the server, it
fetches the timeline history file of the next timeline, and restarts
streaming from the new timeline by issuing a new START_STREAMING command.
When pg_receivexlog crosses a timeline, it leaves the .partial suffix on the
last segment on the old timeline. This helps you to tell apart a partial
segment left in the directory because of a timeline switch, and a completed
segment. If you just follow a single server, it won't make a difference, but
it can be significant in more complicated scenarios where new WAL is still
generated on the old timeline.
This includes two small changes to the streaming replication protocol:
First, when you reach the end of timeline while streaming, the server now
sends the TLI of the next timeline in the server's history to the client.
pg_receivexlog uses that as the next timeline, so that it doesn't need to
parse the timeline history file like a standby server does. Second, when
BASE_BACKUP command sends the begin and end WAL positions, it now also sends
the timeline IDs corresponding the positions.
2013-01-17 19:23:00 +01:00
|
|
|
backup, in a single row with two columns. The first column contains
|
|
|
|
the start position given in XLogRecPtr format, and the second column
|
|
|
|
contains the corresponding timeline ID.
|
2011-01-10 14:03:55 +01:00
|
|
|
</para>
|
|
|
|
<para>
|
2011-02-03 13:46:23 +01:00
|
|
|
The second ordinary result set has one row for each tablespace.
|
2011-01-10 14:03:55 +01:00
|
|
|
The fields in this row are:
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
2015-10-22 04:31:56 +02:00
|
|
|
<term><literal>spcoid</literal> (<type>oid</type>)</term>
|
2011-01-10 14:03:55 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2015-10-22 04:31:56 +02:00
|
|
|
The OID of the tablespace, or null if it's the base
|
2011-01-10 14:03:55 +01:00
|
|
|
directory.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
2015-10-22 04:31:56 +02:00
|
|
|
<term><literal>spclocation</literal> (<type>text</type>)</term>
|
2011-01-10 14:03:55 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2015-10-22 04:31:56 +02:00
|
|
|
The full path of the tablespace directory, or null
|
2011-01-10 14:03:55 +01:00
|
|
|
if it's the base directory.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
2015-10-22 04:31:56 +02:00
|
|
|
<term><literal>size</literal> (<type>int8</type>)</term>
|
2011-01-10 14:03:55 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2019-09-03 11:59:36 +02:00
|
|
|
The approximate size of the tablespace, in kilobytes (1024 bytes),
|
|
|
|
if progress report has been requested; otherwise it's null.
|
2011-01-10 14:03:55 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
Flexible options for BASE_BACKUP.
Previously, BASE_BACKUP used an entirely hard-coded syntax, but that's
hard to extend. Instead, adopt the same kind of syntax we've used for
SQL commands such as VACUUM, ANALYZE, COPY, and EXPLAIN, where it's
not necessary for all of the option names to be parser keywords.
In the new syntax, most of the options now take an optional Boolean
argument. To match our practice in other in places, the options which
the old syntax called NOWAIT and NOVERIFY_CHECKSUMS options are in the
new syntax called WAIT and VERIFY_CHECKUMS, and the default value is
false. In the new syntax, the FAST option has been replaced by a
CHECKSUM option whose value may be 'fast' or 'spread'.
This commit does not remove support for the old syntax. It just adds
the new one as an additional option, and makes pg_basebackup prefer
the new syntax when the server is new enough to support it.
Patch by me, reviewed and tested by Fabien Coelho, Sergei Kornilov,
Fujii Masao, and Tushar Ahuja.
Discussion: http://postgr.es/m/CA+TgmobAczXDRO_Gr2euo_TxgzaH1JxbNxvFx=HYvBinefNH8Q@mail.gmail.com
Discussion: http://postgr.es/m/CA+TgmoZGwR=ZVWFeecncubEyPdwghnvfkkdBe9BLccLSiqdf9Q@mail.gmail.com
2021-10-05 17:50:21 +02:00
|
|
|
|
2011-02-03 13:46:23 +01:00
|
|
|
<para>
|
2022-02-08 17:17:21 +01:00
|
|
|
After the second regular result set, a CopyOutResponse will be sent.
|
|
|
|
The payload of each CopyData message will contain a message in one of
|
Modify pg_basebackup to use a new COPY subprotocol for base backups.
In the new approach, all files across all tablespaces are sent in a
single COPY OUT operation. The CopyData messages are no longer raw
archive content; rather, each message is prefixed with a type byte
that describes its purpose, e.g. 'n' signifies the start of a new
archive and 'd' signifies archive or manifest data. This protocol
is significantly more extensible than the old approach, since we can
later create more message types, though not without concern for
backward compatibility.
The new protocol sends a few things to the client that the old one
did not. First, it sends the name of each archive explicitly, instead
of letting the client compute it. This is intended to make it easier
to write future patches that might send archives in a format other
that tar (e.g. cpio, pax, tar.gz). Second, it sends explicit progress
messages rather than allowing the client to assume that progress is
defined by the number of bytes received. This will help with future
features where the server compresses the data, or sends it someplace
directly rather than transmitting it to the client.
The old protocol is still supported for compatibility with previous
releases. The new protocol is selected by means of a new
TARGET option to the BASE_BACKUP command. Currently, the
only supported target is 'client'. Support for additional
targets will be added in a later commit.
Patch by me. The patch set of which this is a part has had review
and/or testing from Jeevan Ladhe, Tushar Ahuja, Suraj Kharage,
Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+TgmoaYZbz0=Yk797aOJwkGJC-LK3iXn+wzzMx7KdwNpZhS5g@mail.gmail.com
2022-01-18 19:47:26 +01:00
|
|
|
the following formats:
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>new archive (B)</term>
|
|
|
|
<listitem><para><variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte1('n')</term>
|
|
|
|
<listitem><para>
|
|
|
|
Identifes the messaage as indicating the start of a new archive.
|
2022-02-08 17:17:21 +01:00
|
|
|
There will be one archive for the main data directory and one
|
|
|
|
for each additional tablespace; each will use tar format
|
|
|
|
(following the <quote>ustar interchange format</quote> specified
|
|
|
|
in the POSIX 1003.1-2008 standard).
|
Modify pg_basebackup to use a new COPY subprotocol for base backups.
In the new approach, all files across all tablespaces are sent in a
single COPY OUT operation. The CopyData messages are no longer raw
archive content; rather, each message is prefixed with a type byte
that describes its purpose, e.g. 'n' signifies the start of a new
archive and 'd' signifies archive or manifest data. This protocol
is significantly more extensible than the old approach, since we can
later create more message types, though not without concern for
backward compatibility.
The new protocol sends a few things to the client that the old one
did not. First, it sends the name of each archive explicitly, instead
of letting the client compute it. This is intended to make it easier
to write future patches that might send archives in a format other
that tar (e.g. cpio, pax, tar.gz). Second, it sends explicit progress
messages rather than allowing the client to assume that progress is
defined by the number of bytes received. This will help with future
features where the server compresses the data, or sends it someplace
directly rather than transmitting it to the client.
The old protocol is still supported for compatibility with previous
releases. The new protocol is selected by means of a new
TARGET option to the BASE_BACKUP command. Currently, the
only supported target is 'client'. Support for additional
targets will be added in a later commit.
Patch by me. The patch set of which this is a part has had review
and/or testing from Jeevan Ladhe, Tushar Ahuja, Suraj Kharage,
Dipesh Pandit, and Mark Dilger.
Discussion: http://postgr.es/m/CA+TgmoaYZbz0=Yk797aOJwkGJC-LK3iXn+wzzMx7KdwNpZhS5g@mail.gmail.com
2022-01-18 19:47:26 +01:00
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>String</term>
|
|
|
|
<listitem><para>
|
|
|
|
The file name for this archive.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>String</term>
|
|
|
|
<listitem><para>
|
|
|
|
For the main data directory, an empty string. For other
|
|
|
|
tablespaces, the full path to the directory from which this
|
|
|
|
archive was created.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist></para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>manifest (B)</term>
|
|
|
|
<listitem><para><variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte1('m')</term>
|
|
|
|
<listitem><para>
|
|
|
|
Identifes the message as indicating the start of the backup
|
|
|
|
manifest.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist></para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>archive or manifest data (B)</term>
|
|
|
|
<listitem><para><variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte1('d')</term>
|
|
|
|
<listitem><para>
|
|
|
|
Identifes the message as containing archive or manifest data.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte<replaceable>n</replaceable></term>
|
|
|
|
<listitem><para>
|
|
|
|
Data bytes.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist></para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>progress report (B)</term>
|
|
|
|
<listitem><para><variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte1('p')</term>
|
|
|
|
<listitem><para>
|
|
|
|
Identifes the message as a progress report.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>Int64</term>
|
|
|
|
<listitem><para>
|
|
|
|
The number of bytes from the current tablespace for which
|
|
|
|
processing has been completed.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist></para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
After the CopyOutResponse, or all such responses, have been sent, a
|
|
|
|
final ordinary result set will be sent, containing the WAL end position
|
|
|
|
of the backup, in the same format as the start position.
|
2011-02-03 13:46:23 +01:00
|
|
|
</para>
|
|
|
|
|
2011-01-10 14:03:55 +01:00
|
|
|
<para>
|
|
|
|
The tar archive for the data directory and each tablespace will contain
|
|
|
|
all files in the directories, regardless of whether they are
|
2017-10-09 03:44:17 +02:00
|
|
|
<productname>PostgreSQL</productname> files or other files added to the same
|
2011-01-10 14:03:55 +01:00
|
|
|
directory. The only excluded files are:
|
|
|
|
<itemizedlist spacing="compact" mark="bullet">
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
<filename>postmaster.pid</filename>
|
2011-01-10 14:03:55 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
2011-10-19 06:19:43 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
<filename>postmaster.opts</filename>
|
2011-10-19 06:19:43 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
2017-11-07 18:28:35 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-11-08 16:57:27 +01:00
|
|
|
<filename>pg_internal.init</filename> (found in multiple directories)
|
2017-11-07 18:28:35 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
2014-05-15 04:26:49 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2016-09-28 18:00:00 +02:00
|
|
|
Various temporary files and directories created during the operation
|
|
|
|
of the PostgreSQL server, such as any file or directory beginning
|
2018-03-27 15:14:40 +02:00
|
|
|
with <filename>pgsql_tmp</filename> and temporary relations.
|
2014-05-15 04:26:49 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
2018-03-23 17:14:12 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Unlogged relations, except for the init fork which is required to
|
|
|
|
recreate the (empty) unlogged relation on recovery.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
2011-01-10 14:03:55 +01:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
<filename>pg_wal</filename>, including subdirectories. If the backup is run
|
2016-10-20 17:24:37 +02:00
|
|
|
with WAL files included, a synthesized version of <filename>pg_wal</filename> will be
|
2011-01-30 21:30:09 +01:00
|
|
|
included, but it will only contain the files necessary for the
|
|
|
|
backup to work, not the rest of the contents.
|
2011-01-10 14:03:55 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
2014-05-15 04:26:49 +02:00
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
<filename>pg_dynshmem</filename>, <filename>pg_notify</filename>,
|
|
|
|
<filename>pg_replslot</filename>, <filename>pg_serial</filename>,
|
|
|
|
<filename>pg_snapshots</filename>, <filename>pg_stat_tmp</filename>, and
|
|
|
|
<filename>pg_subtrans</filename> are copied as empty directories (even if
|
2016-09-28 18:00:00 +02:00
|
|
|
they are symbolic links).
|
2014-05-15 04:26:49 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Files other than regular files and directories, such as symbolic
|
2016-09-28 18:00:00 +02:00
|
|
|
links (other than for the directories listed above) and special
|
|
|
|
device files, are skipped. (Symbolic links
|
2014-05-15 04:26:49 +02:00
|
|
|
in <filename>pg_tblspc</filename> are maintained.)
|
|
|
|
</para>
|
|
|
|
</listitem>
|
2011-01-10 14:03:55 +01:00
|
|
|
</itemizedlist>
|
2015-10-22 04:31:56 +02:00
|
|
|
Owner, group, and file mode are set if the underlying file system on
|
2011-01-10 14:03:55 +01:00
|
|
|
the server supports it.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2010-06-04 00:17:32 +02:00
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
|
|
|
|
</sect1>
|
|
|
|
|
2017-01-19 18:00:00 +01:00
|
|
|
<sect1 id="protocol-logical-replication">
|
|
|
|
<title>Logical Streaming Replication Protocol</title>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
This section describes the logical replication protocol, which is the message
|
|
|
|
flow started by the <literal>START_REPLICATION</literal>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>SLOT</literal> <replaceable class="parameter">slot_name</replaceable>
|
2017-01-19 18:00:00 +01:00
|
|
|
<literal>LOGICAL</literal> replication command.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The logical streaming replication protocol builds on the primitives of
|
|
|
|
the physical streaming replication protocol.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<sect2 id="protocol-logical-replication-params">
|
|
|
|
<title>Logical Streaming Replication Parameters</title>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The logical replication <literal>START_REPLICATION</literal> command
|
|
|
|
accepts following parameters:
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
proto_version
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
Protocol version. Currently versions <literal>1</literal>, <literal>2</literal>,
|
|
|
|
and <literal>3</literal> are supported.
|
|
|
|
</para>
|
|
|
|
<para>
|
|
|
|
Version <literal>2</literal> is supported only for server version 14
|
|
|
|
and above, and it allows streaming of large in-progress transactions.
|
|
|
|
</para>
|
|
|
|
<para>
|
|
|
|
Version <literal>3</literal> is supported only for server version 15
|
|
|
|
and above, and it allows streaming of two-phase transactions.
|
|
|
|
</para>
|
2017-01-19 18:00:00 +01:00
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
publication_names
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Comma separated list of publication names for which to subscribe
|
|
|
|
(receive changes). The individual publication names are treated
|
|
|
|
as standard objects names and can be quoted the same as needed.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</sect2>
|
|
|
|
|
|
|
|
<sect2 id="protocol-logical-messages">
|
|
|
|
<title>Logical Replication Protocol Messages</title>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The individual protocol messages are discussed in the following
|
2017-03-19 04:43:47 +01:00
|
|
|
subsections. Individual messages are described in
|
2017-11-23 15:39:47 +01:00
|
|
|
<xref linkend="protocol-logicalrep-message-formats"/>.
|
2017-01-19 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
All top-level protocol messages begin with a message type byte.
|
|
|
|
While represented in code as a character, this is a signed byte with no
|
|
|
|
associated encoding.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Since the streaming replication protocol supplies a message length there
|
|
|
|
is no need for top-level protocol messages to embed a length in their
|
|
|
|
header.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
</sect2>
|
|
|
|
|
|
|
|
<sect2 id="protocol-logical-messages-flow">
|
|
|
|
<title>Logical Replication Protocol Message Flow</title>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
With the exception of the <literal>START_REPLICATION</literal> command and
|
|
|
|
the replay progress messages, all information flows only from the backend
|
|
|
|
to the frontend.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The logical replication protocol sends individual transactions one by one.
|
|
|
|
This means that all messages between a pair of Begin and Commit messages
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
belong to the same transaction. Similarly, all messages between a pair of
|
|
|
|
Begin Prepare and Prepare messages belong to the same transaction.
|
|
|
|
It also sends changes of large in-progress transactions between a pair of
|
2021-08-02 17:32:17 +02:00
|
|
|
Stream Start and Stream Stop messages. The last stream of such a transaction
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
contains a Stream Commit or Stream Abort message.
|
2017-01-19 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Every sent transaction contains zero or more DML messages (Insert,
|
|
|
|
Update, Delete). In case of a cascaded setup it can also contain Origin
|
2020-10-19 18:28:54 +02:00
|
|
|
messages. The origin message indicates that the transaction originated on
|
2017-01-19 18:00:00 +01:00
|
|
|
different replication node. Since a replication node in the scope of logical
|
|
|
|
replication protocol can be pretty much anything, the only identifier
|
|
|
|
is the origin name. It's downstream's responsibility to handle this as
|
|
|
|
needed (if needed). The Origin message is always sent before any DML
|
|
|
|
messages in the transaction.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2021-11-10 19:12:58 +01:00
|
|
|
Every DML message contains a relation OID, identifying the publisher's
|
|
|
|
relation that was acted on. Before the first DML message for a given
|
|
|
|
relation OID, a Relation message will be sent, describing the schema of
|
|
|
|
that relation. Subsequently, a new Relation message will be sent if
|
|
|
|
the relation's definition has changed since the last Relation message
|
|
|
|
was sent for it. (The protocol assumes that the client is capable of
|
|
|
|
remembering this metadata for as many relations as needed.)
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Relation messages identify column types by their OIDs. In the case
|
|
|
|
of a built-in type, it is assumed that the client can look up that
|
|
|
|
type OID locally, so no additional data is needed. For a non-built-in
|
|
|
|
type OID, a Type message will be sent before the Relation message,
|
|
|
|
to provide the type name associated with that OID. Thus, a client that
|
|
|
|
needs to specifically identify the types of relation columns should
|
|
|
|
cache the contents of Type messages, and first consult that cache to
|
|
|
|
see if the type OID is defined there. If not, look up the type OID
|
|
|
|
locally.
|
2017-01-19 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
</sect2>
|
|
|
|
</sect1>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<sect1 id="protocol-message-types">
|
|
|
|
<title>Message Data Types</title>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
This section describes the base data types used in messages.
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int<replaceable>n</replaceable>(<replaceable>i</replaceable>)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2012-09-01 18:04:44 +02:00
|
|
|
An <replaceable>n</replaceable>-bit integer in network byte
|
2005-06-22 17:19:43 +02:00
|
|
|
order (most significant byte first).
|
2005-01-23 01:30:59 +01:00
|
|
|
If <replaceable>i</replaceable> is specified it
|
2003-04-16 00:51:18 +02:00
|
|
|
is the exact value that will appear, otherwise the value
|
2005-06-22 17:19:43 +02:00
|
|
|
is variable. Eg. Int16, Int32(42).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int<replaceable>n</replaceable>[<replaceable>k</replaceable>]
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2005-06-22 17:19:43 +02:00
|
|
|
An array of <replaceable>k</replaceable>
|
2005-01-23 01:30:59 +01:00
|
|
|
<replaceable>n</replaceable>-bit integers, each in network
|
2005-06-22 17:19:43 +02:00
|
|
|
byte order. The array length <replaceable>k</replaceable>
|
|
|
|
is always determined by an earlier field in the message.
|
|
|
|
Eg. Int16[M].
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String(<replaceable>s</replaceable>)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
A null-terminated string (C-style string). There is no
|
2005-06-22 17:19:43 +02:00
|
|
|
specific length limitation on strings.
|
2005-01-23 01:30:59 +01:00
|
|
|
If <replaceable>s</replaceable> is specified it is the exact
|
2005-06-22 17:19:43 +02:00
|
|
|
value that will appear, otherwise the value is variable.
|
2001-06-23 01:27:48 +02:00
|
|
|
Eg. String, String("user").
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
2010-02-16 23:34:57 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<note>
|
|
|
|
<para>
|
|
|
|
<emphasis>There is no predefined limit</emphasis> on the length of a string
|
1999-09-12 00:02:51 +02:00
|
|
|
that can be returned by the backend. Good coding strategy for a frontend
|
|
|
|
is to use an expandable buffer so that anything that fits in memory can be
|
|
|
|
accepted. If that's not feasible, read the full string and discard trailing
|
|
|
|
characters that don't fit into your fixed-size buffer.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</note>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>(<replaceable>c</replaceable>)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Exactly <replaceable>n</replaceable> bytes. If the field
|
2005-06-22 17:19:43 +02:00
|
|
|
width <replaceable>n</replaceable> is not a constant, it is
|
|
|
|
always determinable from an earlier field in the message.
|
2005-01-23 01:30:59 +01:00
|
|
|
If <replaceable>c</replaceable> is specified it is the exact
|
2003-04-16 00:51:18 +02:00
|
|
|
value. Eg. Byte2, Byte1('\n').
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</variablelist>
|
|
|
|
</para>
|
1998-12-29 03:24:47 +01:00
|
|
|
</sect1>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<sect1 id="protocol-message-formats">
|
|
|
|
<title>Message Formats</title>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
This section describes the detailed format of each message. Each is marked to
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
indicate that it can be sent by a frontend (F), a backend (B), or both
|
2003-04-16 00:51:18 +02:00
|
|
|
(F & B).
|
|
|
|
Notice that although each message includes a byte count at the beginning,
|
|
|
|
the message format is defined so that the message end can be found without
|
2003-04-19 02:02:30 +02:00
|
|
|
reference to the byte count. This aids validity checking. (The CopyData
|
|
|
|
message is an exception, because it forms part of a data stream; the contents
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
of any individual CopyData message cannot be interpretable on their own.)
|
1998-12-29 03:24:47 +01:00
|
|
|
</para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationOk">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
AuthenticationOk (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('R')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as an authentication request.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(8)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(0)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Specifies that the authentication was successful.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationKerberosV5">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
1998-03-01 09:16:16 +01:00
|
|
|
AuthenticationKerberosV5 (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-06-23 01:27:48 +02:00
|
|
|
Byte1('R')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
Identifies the message as an authentication request.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(8)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-06-23 01:27:48 +02:00
|
|
|
Int32(2)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
Specifies that Kerberos V5 authentication is required.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationCleartextPassword">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2001-09-21 22:31:49 +02:00
|
|
|
AuthenticationCleartextPassword (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-06-23 01:27:48 +02:00
|
|
|
Byte1('R')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
Identifies the message as an authentication request.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(8)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(3)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-11-01 02:56:29 +01:00
|
|
|
Specifies that a clear-text password is required.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationMD5Password">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2001-09-21 22:31:49 +02:00
|
|
|
AuthenticationMD5Password (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-09-21 22:31:49 +02:00
|
|
|
Byte1('R')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-09-21 22:31:49 +02:00
|
|
|
Identifies the message as an authentication request.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(12)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-09-21 22:31:49 +02:00
|
|
|
Int32(5)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-09-21 22:31:49 +02:00
|
|
|
Specifies that an MD5-encrypted password is required.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-09-21 22:31:49 +02:00
|
|
|
Byte4
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-09-21 22:31:49 +02:00
|
|
|
The salt to use when encrypting the password.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationSCMCredential">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2001-09-21 22:31:49 +02:00
|
|
|
AuthenticationSCMCredential (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-09-21 22:31:49 +02:00
|
|
|
Byte1('R')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-09-21 22:31:49 +02:00
|
|
|
Identifies the message as an authentication request.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(8)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-09-21 22:31:49 +02:00
|
|
|
Int32(6)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-09-21 22:31:49 +02:00
|
|
|
Specifies that an SCM credentials message is required.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationGSS">
|
2007-07-18 14:00:47 +02:00
|
|
|
<term>
|
|
|
|
AuthenticationGSS (B)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('R')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an authentication request.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32(8)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32(7)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Specifies that GSSAPI authentication is required.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationGSSContinue">
|
2007-12-03 14:40:11 +01:00
|
|
|
<term>
|
2020-10-23 13:01:39 +02:00
|
|
|
AuthenticationGSSContinue (B)
|
2007-12-03 14:40:11 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('R')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an authentication request.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2020-10-23 13:01:39 +02:00
|
|
|
Int32
|
2007-12-03 14:40:11 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2020-10-23 13:01:39 +02:00
|
|
|
Int32(8)
|
2007-12-03 14:40:11 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2020-10-23 13:01:39 +02:00
|
|
|
Specifies that this message contains GSSAPI or SSPI data.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
GSSAPI or SSPI authentication data.
|
2007-12-03 14:40:11 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationSSPI">
|
2007-07-18 14:00:47 +02:00
|
|
|
<term>
|
2020-10-23 13:01:39 +02:00
|
|
|
AuthenticationSSPI (B)
|
2007-07-18 14:00:47 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('R')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an authentication request.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32(8)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2020-10-23 13:01:39 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2007-07-18 14:00:47 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2020-10-23 13:01:39 +02:00
|
|
|
Int32(9)
|
2007-07-18 14:00:47 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2020-10-23 13:01:39 +02:00
|
|
|
Specifies that SSPI authentication is required.
|
2007-07-18 14:00:47 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationSASL">
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
<term>
|
|
|
|
AuthenticationSASL (B)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('R')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an authentication request.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32(10)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
Specifies that SASL authentication is required.
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</variablelist>
|
|
|
|
The message body is a list of SASL authentication mechanisms, in the
|
|
|
|
server's order of preference. A zero byte is required as terminator after
|
|
|
|
the last authentication mechanism name. For each mechanism, there is the
|
|
|
|
following:
|
|
|
|
<variablelist>
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Name of a SASL authentication mechanism.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationSASLContinue">
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
<term>
|
|
|
|
AuthenticationSASLContinue (B)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('R')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an authentication request.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32(11)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
Specifies that this message contains a SASL challenge.
|
Support SCRAM-SHA-256 authentication (RFC 5802 and 7677).
This introduces a new generic SASL authentication method, similar to the
GSS and SSPI methods. The server first tells the client which SASL
authentication mechanism to use, and then the mechanism-specific SASL
messages are exchanged in AuthenticationSASLcontinue and PasswordMessage
messages. Only SCRAM-SHA-256 is supported at the moment, but this allows
adding more SASL mechanisms in the future, without changing the overall
protocol.
Support for channel binding, aka SCRAM-SHA-256-PLUS is left for later.
The SASLPrep algorithm, for pre-processing the password, is not yet
implemented. That could cause trouble, if you use a password with
non-ASCII characters, and a client library that does implement SASLprep.
That will hopefully be added later.
Authorization identities, as specified in the SCRAM-SHA-256 specification,
are ignored. SET SESSION AUTHORIZATION provides more or less the same
functionality, anyway.
If a user doesn't exist, perform a "mock" authentication, by constructing
an authentic-looking challenge on the fly. The challenge is derived from
a new system-wide random value, "mock authentication nonce", which is
created at initdb, and stored in the control file. We go through these
motions, in order to not give away the information on whether the user
exists, to unauthenticated users.
Bumps PG_CONTROL_VERSION, because of the new field in control file.
Patch by Michael Paquier and Heikki Linnakangas, reviewed at different
stages by Robert Haas, Stephen Frost, David Steele, Aleksander Alekseev,
and many others.
Discussion: https://www.postgresql.org/message-id/CAB7nPqRbR3GmFYdedCAhzukfKrgBLTLtMvENOmPrVWREsZkF8g%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/CAB7nPqSMXU35g%3DW9X74HVeQp0uvgJxvYOuA4A-A3M%2B0wfEBv-w%40mail.gmail.com
Discussion: https://www.postgresql.org/message-id/55192AFE.6080106@iki.fi
2017-03-07 13:25:40 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
SASL data, specific to the SASL mechanism being used.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2007-07-18 14:00:47 +02:00
|
|
|
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-AuthenticationSASLFinal">
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
<term>
|
|
|
|
AuthenticationSASLFinal (B)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('R')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an authentication request.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32(12)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Specifies that SASL authentication has completed.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
SASL outcome "additional data", specific to the SASL mechanism
|
|
|
|
being used.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-BackendKeyData">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
1998-07-09 05:29:11 +02:00
|
|
|
BackendKeyData (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-07-09 05:29:11 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-06-23 01:27:48 +02:00
|
|
|
Byte1('K')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
Identifies the message as cancellation key data.
|
|
|
|
The frontend must save these values if it wishes to be
|
|
|
|
able to issue CancelRequest messages later.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(12)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-06-23 01:27:48 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
The process ID of this backend.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-06-23 01:27:48 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-06-23 01:27:48 +02:00
|
|
|
The secret key of this backend.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-07-09 05:29:11 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-Bind">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Bind (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-07-09 05:29:11 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('B')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Bind command.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The name of the destination portal
|
2005-06-22 17:19:43 +02:00
|
|
|
(an empty string selects the unnamed portal).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The name of the source prepared statement
|
2005-06-22 17:19:43 +02:00
|
|
|
(an empty string selects the unnamed prepared statement).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2005-06-22 17:19:43 +02:00
|
|
|
The number of parameter format codes that follow
|
2017-10-09 03:44:17 +02:00
|
|
|
(denoted <replaceable>C</replaceable> below).
|
2005-06-22 17:19:43 +02:00
|
|
|
This can be zero to indicate that there are no parameters
|
|
|
|
or that the parameters all use the default format (text);
|
|
|
|
or one, in which case the specified format code is applied
|
|
|
|
to all parameters; or it can equal the actual number of
|
|
|
|
parameters.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
Int16[<replaceable>C</replaceable>]
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The parameter format codes. Each must presently be
|
2005-06-22 17:19:43 +02:00
|
|
|
zero (text) or one (binary).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The number of parameter values that follow (possibly zero).
|
2005-06-22 17:19:43 +02:00
|
|
|
This must match the number of parameters needed by the query.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-05-07 23:46:15 +02:00
|
|
|
Next, the following pair of fields appear for each parameter:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The length of the parameter value, in bytes (this count
|
2005-06-22 17:19:43 +02:00
|
|
|
does not include itself). Can be zero.
|
|
|
|
As a special case, -1 indicates a NULL parameter value.
|
|
|
|
No value bytes follow in the NULL case.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The value of the parameter, in the format indicated by the
|
2005-06-22 17:19:43 +02:00
|
|
|
associated format code.
|
2005-01-23 01:30:59 +01:00
|
|
|
<replaceable>n</replaceable> is the above length.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-05-07 23:46:15 +02:00
|
|
|
After the last parameter, the following fields appear:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2005-06-22 17:19:43 +02:00
|
|
|
The number of result-column format codes that follow
|
2017-10-09 03:44:17 +02:00
|
|
|
(denoted <replaceable>R</replaceable> below).
|
2005-06-22 17:19:43 +02:00
|
|
|
This can be zero to indicate that there are no result columns
|
|
|
|
or that the result columns should all use the default format
|
2010-02-16 23:34:57 +01:00
|
|
|
(text);
|
2005-06-22 17:19:43 +02:00
|
|
|
or one, in which case the specified format code is applied
|
|
|
|
to all result columns (if any); or it can equal the actual
|
|
|
|
number of result columns of the query.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
Int16[<replaceable>R</replaceable>]
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The result-column format codes. Each must presently be
|
2005-06-22 17:19:43 +02:00
|
|
|
zero (text) or one (binary).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-BindComplete">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
BindComplete (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('2')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Bind-complete indicator.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-CancelRequest">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
CancelRequest (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(16)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(80877102)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The cancel request code. The value is chosen to contain
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>1234</literal> in the most significant 16 bits, and <literal>5678</literal> in the
|
2016-05-04 03:06:25 +02:00
|
|
|
least significant 16 bits. (To avoid confusion, this code
|
2003-04-16 00:51:18 +02:00
|
|
|
must not be the same as any protocol version number.)
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The process ID of the target backend.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The secret key for the target backend.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-Close">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Close (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('C')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Close command.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-06-23 01:27:48 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
'<literal>S</literal>' to close a prepared statement; or
|
|
|
|
'<literal>P</literal>' to close a portal.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The name of the prepared statement or portal to close
|
2005-06-22 17:19:43 +02:00
|
|
|
(an empty string selects the unnamed prepared statement
|
|
|
|
or portal).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
1998-12-29 03:24:47 +01:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-CloseComplete">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-05-05 02:44:56 +02:00
|
|
|
CloseComplete (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-05 02:44:56 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-05 02:44:56 +02:00
|
|
|
Byte1('3')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-05 02:44:56 +02:00
|
|
|
Identifies the message as a Close-complete indicator.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-05 02:44:56 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-05 02:44:56 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-05-05 02:44:56 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-05-05 02:44:56 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-CommandComplete">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
CommandComplete (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('C')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a command-completed response.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2006-03-03 20:54:10 +01:00
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The command tag. This is usually a single
|
|
|
|
word that identifies which SQL command was completed.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
For an <command>INSERT</command> command, the tag is
|
|
|
|
<literal>INSERT <replaceable>oid</replaceable>
|
|
|
|
<replaceable>rows</replaceable></literal>, where
|
|
|
|
<replaceable>rows</replaceable> is the number of rows
|
Remove WITH OIDS support, change oid catalog column visibility.
Previously tables declared WITH OIDS, including a significant fraction
of the catalog tables, stored the oid column not as a normal column,
but as part of the tuple header.
This special column was not shown by default, which was somewhat odd,
as it's often (consider e.g. pg_class.oid) one of the more important
parts of a row. Neither pg_dump nor COPY included the contents of the
oid column by default.
The fact that the oid column was not an ordinary column necessitated a
significant amount of special case code to support oid columns. That
already was painful for the existing, but upcoming work aiming to make
table storage pluggable, would have required expanding and duplicating
that "specialness" significantly.
WITH OIDS has been deprecated since 2005 (commit ff02d0a05280e0).
Remove it.
Removing includes:
- CREATE TABLE and ALTER TABLE syntax for declaring the table to be
WITH OIDS has been removed (WITH (oids[ = true]) will error out)
- pg_dump does not support dumping tables declared WITH OIDS and will
issue a warning when dumping one (and ignore the oid column).
- restoring an pg_dump archive with pg_restore will warn when
restoring a table with oid contents (and ignore the oid column)
- COPY will refuse to load binary dump that includes oids.
- pg_upgrade will error out when encountering tables declared WITH
OIDS, they have to be altered to remove the oid column first.
- Functionality to access the oid of the last inserted row (like
plpgsql's RESULT_OID, spi's SPI_lastoid, ...) has been removed.
The syntax for declaring a table WITHOUT OIDS (or WITH (oids = false)
for CREATE TABLE) is still supported. While that requires a bit of
support code, it seems unnecessary to break applications / dumps that
do not use oids, and are explicit about not using them.
The biggest user of WITH OID columns was postgres' catalog. This
commit changes all 'magic' oid columns to be columns that are normally
declared and stored. To reduce unnecessary query breakage all the
newly added columns are still named 'oid', even if a table's column
naming scheme would indicate 'reloid' or such. This obviously
requires adapting a lot code, mostly replacing oid access via
HeapTupleGetOid() with access to the underlying Form_pg_*->oid column.
The bootstrap process now assigns oids for all oid columns in
genbki.pl that do not have an explicit value (starting at the largest
oid previously used), only oids assigned later by oids will be above
FirstBootstrapObjectId. As the oid column now is a normal column the
special bootstrap syntax for oids has been removed.
Oids are not automatically assigned during insertion anymore, all
backend code explicitly assigns oids with GetNewOidWithIndex(). For
the rare case that insertions into the catalog via SQL are called for
the new pg_nextoid() function can be used (which only works on catalog
tables).
The fact that oid columns on system tables are now normal columns
means that they will be included in the set of columns expanded
by * (i.e. SELECT * FROM pg_class will now include the table's oid,
previously it did not). It'd not technically be hard to hide oid
column by default, but that'd mean confusing behavior would either
have to be carried forward forever, or it'd cause breakage down the
line.
While it's not unlikely that further adjustments are needed, the
scope/invasiveness of the patch makes it worthwhile to get merge this
now. It's painful to maintain externally, too complicated to commit
after the code code freeze, and a dependency of a number of other
patches.
Catversion bump, for obvious reasons.
Author: Andres Freund, with contributions by John Naylor
Discussion: https://postgr.es/m/20180930034810.ywp2c7awz7opzcfr@alap3.anarazel.de
2018-11-21 00:36:57 +01:00
|
|
|
inserted. <replaceable>oid</replaceable> used to be the object ID
|
|
|
|
of the inserted row if <replaceable>rows</replaceable> was 1
|
|
|
|
and the target table had OIDs, but OIDs system columns are
|
|
|
|
not supported anymore; therefore <replaceable>oid</replaceable>
|
|
|
|
is always 0.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
For a <command>DELETE</command> command, the tag is
|
2005-06-22 17:19:43 +02:00
|
|
|
<literal>DELETE <replaceable>rows</replaceable></literal> where
|
2005-01-23 01:30:59 +01:00
|
|
|
<replaceable>rows</replaceable> is the number of rows deleted.
|
|
|
|
</para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
For an <command>UPDATE</command> command, the tag is
|
2005-06-22 17:19:43 +02:00
|
|
|
<literal>UPDATE <replaceable>rows</replaceable></literal> where
|
2005-01-23 01:30:59 +01:00
|
|
|
<replaceable>rows</replaceable> is the number of rows updated.
|
|
|
|
</para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2010-02-16 21:58:14 +01:00
|
|
|
<para>
|
|
|
|
For a <command>SELECT</command> or <command>CREATE TABLE AS</command>
|
|
|
|
command, the tag is <literal>SELECT <replaceable>rows</replaceable></literal>
|
|
|
|
where <replaceable>rows</replaceable> is the number of rows retrieved.
|
|
|
|
</para>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
|
|
|
For a <command>MOVE</command> command, the tag is
|
|
|
|
<literal>MOVE <replaceable>rows</replaceable></literal> where
|
|
|
|
<replaceable>rows</replaceable> is the number of rows the
|
|
|
|
cursor's position has been changed by.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
For a <command>FETCH</command> command, the tag is
|
|
|
|
<literal>FETCH <replaceable>rows</replaceable></literal> where
|
|
|
|
<replaceable>rows</replaceable> is the number of rows that
|
|
|
|
have been retrieved from the cursor.
|
2006-03-03 20:54:10 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
For a <command>COPY</command> command, the tag is
|
|
|
|
<literal>COPY <replaceable>rows</replaceable></literal> where
|
|
|
|
<replaceable>rows</replaceable> is the number of rows copied.
|
|
|
|
(Note: the row count appears only in
|
|
|
|
<productname>PostgreSQL</productname> 8.2 and later.)
|
|
|
|
</para>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-CopyData">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-19 02:02:30 +02:00
|
|
|
CopyData (F & B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('d')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
Identifies the message as <command>COPY</command> data.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-11-01 02:56:29 +01:00
|
|
|
Data that forms part of a <command>COPY</command> data stream. Messages sent
|
2005-06-22 17:19:43 +02:00
|
|
|
from the backend will always correspond to single data rows,
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
but messages sent by frontends might divide the data stream
|
2005-06-22 17:19:43 +02:00
|
|
|
arbitrarily.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-CopyDone">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
CopyDone (F & B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('c')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
Identifies the message as a <command>COPY</command>-complete indicator.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-CopyFail">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
CopyFail (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('f')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
Identifies the message as a <command>COPY</command>-failure indicator.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
An error message to report as the cause of failure.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-CopyInResponse">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
CopyInResponse (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('G')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Start Copy In response.
|
|
|
|
The frontend must now send copy-in data (if not
|
2005-06-22 17:19:43 +02:00
|
|
|
prepared to do so, send a CopyFail message).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 22:53:38 +02:00
|
|
|
Int8
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
0 indicates the overall <command>COPY</command> format is textual (rows
|
2005-06-22 17:19:43 +02:00
|
|
|
separated by newlines, columns separated by separator
|
|
|
|
characters, etc).
|
|
|
|
1 indicates the overall copy format is binary (similar
|
|
|
|
to DataRow format).
|
2017-11-23 15:39:47 +01:00
|
|
|
See <xref linkend="sql-copy"/>
|
2005-06-22 17:19:43 +02:00
|
|
|
for more information.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2005-06-22 17:19:43 +02:00
|
|
|
The number of columns in the data to be copied
|
2017-10-09 03:44:17 +02:00
|
|
|
(denoted <replaceable>N</replaceable> below).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
Int16[<replaceable>N</replaceable>]
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The format codes to be used for each column.
|
2005-06-22 17:19:43 +02:00
|
|
|
Each must presently be zero (text) or one (binary).
|
|
|
|
All must be zero if the overall copy format is textual.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-CopyOutResponse">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
CopyOutResponse (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('H')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Start Copy Out response.
|
|
|
|
This message will be followed by copy-out data.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
2010-12-11 15:27:37 +01:00
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int8
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
0 indicates the overall <command>COPY</command> format
|
|
|
|
is textual (rows separated by newlines, columns
|
|
|
|
separated by separator characters, etc). 1 indicates
|
|
|
|
the overall copy format is binary (similar to DataRow
|
2017-11-23 15:39:47 +01:00
|
|
|
format). See <xref linkend="sql-copy"/> for more information.
|
2010-12-11 15:27:37 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int16
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The number of columns in the data to be copied
|
2017-10-09 03:44:17 +02:00
|
|
|
(denoted <replaceable>N</replaceable> below).
|
2010-12-11 15:27:37 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
Int16[<replaceable>N</replaceable>]
|
2010-12-11 15:27:37 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The format codes to be used for each column.
|
|
|
|
Each must presently be zero (text) or one (binary).
|
|
|
|
All must be zero if the overall copy format is textual.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-CopyBothResponse">
|
2010-12-11 15:27:37 +01:00
|
|
|
<term>
|
|
|
|
CopyBothResponse (B)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('W')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a Start Copy Both response.
|
|
|
|
This message is used only for Streaming Replication.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
2005-01-23 01:30:59 +01:00
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 22:53:38 +02:00
|
|
|
Int8
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
0 indicates the overall <command>COPY</command> format
|
|
|
|
is textual (rows separated by newlines, columns
|
|
|
|
separated by separator characters, etc). 1 indicates
|
|
|
|
the overall copy format is binary (similar to DataRow
|
2017-11-23 15:39:47 +01:00
|
|
|
format). See <xref linkend="sql-copy"/> for more information.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2005-06-22 17:19:43 +02:00
|
|
|
The number of columns in the data to be copied
|
2017-10-09 03:44:17 +02:00
|
|
|
(denoted <replaceable>N</replaceable> below).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
Int16[<replaceable>N</replaceable>]
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The format codes to be used for each column.
|
2005-06-22 17:19:43 +02:00
|
|
|
Each must presently be zero (text) or one (binary).
|
|
|
|
All must be zero if the overall copy format is textual.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-DataRow">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
DataRow (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('D')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
Identifies the message as a data row.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The number of column values that follow (possibly zero).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-05-07 23:46:15 +02:00
|
|
|
Next, the following pair of fields appear for each column:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The length of the column value, in bytes (this count
|
2005-06-22 17:19:43 +02:00
|
|
|
does not include itself). Can be zero.
|
|
|
|
As a special case, -1 indicates a NULL column value.
|
|
|
|
No value bytes follow in the NULL case.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The value of the column, in the format indicated by the
|
2005-06-22 17:19:43 +02:00
|
|
|
associated format code.
|
2005-01-23 01:30:59 +01:00
|
|
|
<replaceable>n</replaceable> is the above length.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-Describe">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Describe (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('D')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Describe command.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
'<literal>S</literal>' to describe a prepared statement; or
|
|
|
|
'<literal>P</literal>' to describe a portal.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The name of the prepared statement or portal to describe
|
2005-06-22 17:19:43 +02:00
|
|
|
(an empty string selects the unnamed prepared statement
|
|
|
|
or portal).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-EmptyQueryResponse">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
EmptyQueryResponse (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('I')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a response to an empty query string.
|
2005-06-22 17:19:43 +02:00
|
|
|
(This substitutes for CommandComplete.)
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 22:53:38 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-ErrorResponse">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
ErrorResponse (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('E')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as an error.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
The message body consists of one or more identified fields,
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
followed by a zero byte as a terminator. Fields can appear in
|
2005-06-22 17:19:43 +02:00
|
|
|
any order. For each field there is the following:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
A code identifying the field type; if zero, this is
|
2005-06-22 17:19:43 +02:00
|
|
|
the message terminator and no string follows.
|
|
|
|
The presently defined field types are listed in
|
2017-11-23 15:39:47 +01:00
|
|
|
<xref linkend="protocol-error-fields"/>.
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
Since more field types might be added in future,
|
2005-06-22 17:19:43 +02:00
|
|
|
frontends should silently ignore fields of unrecognized
|
|
|
|
type.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The field value.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-Execute">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Execute (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('E')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as an Execute command.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The name of the portal to execute
|
2005-06-22 17:19:43 +02:00
|
|
|
(an empty string selects the unnamed portal).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Maximum number of rows to return, if portal contains
|
2005-06-22 17:19:43 +02:00
|
|
|
a query that returns rows (ignored otherwise). Zero
|
2017-10-09 03:44:17 +02:00
|
|
|
denotes <quote>no limit</quote>.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-Flush">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Flush (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('H')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Flush command.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-FunctionCall">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
FunctionCall (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('F')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a function call.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Specifies the object ID of the function to call.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2005-06-22 17:19:43 +02:00
|
|
|
The number of argument format codes that follow
|
2017-10-09 03:44:17 +02:00
|
|
|
(denoted <replaceable>C</replaceable> below).
|
2005-06-22 17:19:43 +02:00
|
|
|
This can be zero to indicate that there are no arguments
|
|
|
|
or that the arguments all use the default format (text);
|
|
|
|
or one, in which case the specified format code is applied
|
|
|
|
to all arguments; or it can equal the actual number of
|
|
|
|
arguments.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
Int16[<replaceable>C</replaceable>]
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The argument format codes. Each must presently be
|
2005-06-22 17:19:43 +02:00
|
|
|
zero (text) or one (binary).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
Specifies the number of arguments being supplied to the
|
|
|
|
function.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-05-07 23:46:15 +02:00
|
|
|
Next, the following pair of fields appear for each argument:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The length of the argument value, in bytes (this count
|
2005-06-22 17:19:43 +02:00
|
|
|
does not include itself). Can be zero.
|
|
|
|
As a special case, -1 indicates a NULL argument value.
|
|
|
|
No value bytes follow in the NULL case.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The value of the argument, in the format indicated by the
|
2005-06-22 17:19:43 +02:00
|
|
|
associated format code.
|
2005-01-23 01:30:59 +01:00
|
|
|
<replaceable>n</replaceable> is the above length.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-05-07 23:46:15 +02:00
|
|
|
After the last argument, the following field appears:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The format code for the function result. Must presently be
|
2005-06-22 17:19:43 +02:00
|
|
|
zero (text) or one (binary).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-FunctionCallResponse">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
FunctionCallResponse (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('V')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a function call result.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The length of the function result value, in bytes (this count
|
2005-06-22 17:19:43 +02:00
|
|
|
does not include itself). Can be zero.
|
|
|
|
As a special case, -1 indicates a NULL function result.
|
|
|
|
No value bytes follow in the NULL case.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The value of the function result, in the format indicated by
|
2005-06-22 17:19:43 +02:00
|
|
|
the associated format code.
|
2005-01-23 01:30:59 +01:00
|
|
|
<replaceable>n</replaceable> is the above length.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-GSSENCRequest">
|
2020-10-27 08:43:35 +01:00
|
|
|
<term>
|
|
|
|
GSSENCRequest (F)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32(8)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32(80877104)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The <acronym>GSSAPI</acronym> Encryption request code. The value is chosen to contain
|
|
|
|
<literal>1234</literal> in the most significant 16 bits, and <literal>5680</literal> in the
|
|
|
|
least significant 16 bits. (To avoid confusion, this code
|
|
|
|
must not be the same as any protocol version number.)
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-GSSResponse">
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
<term>
|
|
|
|
GSSResponse (F)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('p')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a GSSAPI or SSPI response. Note that
|
|
|
|
this is also used for SASL and password response messages.
|
|
|
|
The exact message type can be deduced from the context.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
GSSAPI/SSPI specific message data.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-NegotiateProtocolVersion">
|
2017-11-21 19:56:24 +01:00
|
|
|
<term>
|
|
|
|
NegotiateProtocolVersion (B)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('v')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a protocol version negotiation
|
|
|
|
message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Newest minor protocol version supported by the server
|
|
|
|
for the major protocol version requested by the client.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Number of protocol options not recognized by the server.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
Then, for protocol option not recognized by the server, there
|
|
|
|
is the following:
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The option name.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-NoData">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
NoData (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('n')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a no-data indicator.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-NoticeResponse">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
NoticeResponse (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('N')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a notice.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
The message body consists of one or more identified fields,
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
followed by a zero byte as a terminator. Fields can appear in
|
2005-06-22 17:19:43 +02:00
|
|
|
any order. For each field there is the following:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
A code identifying the field type; if zero, this is
|
2005-06-22 17:19:43 +02:00
|
|
|
the message terminator and no string follows.
|
|
|
|
The presently defined field types are listed in
|
2017-11-23 15:39:47 +01:00
|
|
|
<xref linkend="protocol-error-fields"/>.
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
Since more field types might be added in future,
|
2005-06-22 17:19:43 +02:00
|
|
|
frontends should silently ignore fields of unrecognized
|
|
|
|
type.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The field value.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-NotificationResponse">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
NotificationResponse (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('A')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a notification response.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The process ID of the notifying backend process.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2010-02-16 23:34:57 +01:00
|
|
|
The name of the channel that the notify has been raised on.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The <quote>payload</quote> string passed from the notifying process.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-ParameterDescription">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
ParameterDescription (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('t')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a parameter description.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The number of parameters used by the statement
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
(can be zero).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
Then, for each parameter, there is the following:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-11-01 02:56:29 +01:00
|
|
|
Specifies the object ID of the parameter data type.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-ParameterStatus">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
ParameterStatus (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('S')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a run-time parameter status report.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The name of the run-time parameter being reported.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The current value of the parameter.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-Parse">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Parse (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('P')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Parse command.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The name of the destination prepared statement
|
2005-06-22 17:19:43 +02:00
|
|
|
(an empty string selects the unnamed prepared statement).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The query string to be parsed.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-11-01 02:56:29 +01:00
|
|
|
The number of parameter data types specified
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
(can be zero). Note that this is not an indication of
|
2005-06-22 17:19:43 +02:00
|
|
|
the number of parameters that might appear in the
|
|
|
|
query string, only the number that the frontend wants to
|
|
|
|
prespecify types for.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
Then, for each parameter, there is the following:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-11-01 02:56:29 +01:00
|
|
|
Specifies the object ID of the parameter data type.
|
2005-06-22 17:19:43 +02:00
|
|
|
Placing a zero here is equivalent to leaving the type
|
|
|
|
unspecified.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-ParseComplete">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
ParseComplete (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('1')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Parse-complete indicator.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-PasswordMessage">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
PasswordMessage (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('p')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2007-07-18 14:00:47 +02:00
|
|
|
Identifies the message as a password response. Note that
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
this is also used for GSSAPI, SSPI and SASL response messages.
|
|
|
|
The exact message type can be deduced from the context.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The password (encrypted, if requested).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-PortalSuspended">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
PortalSuspended (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('s')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a portal-suspended indicator.
|
2005-06-22 17:19:43 +02:00
|
|
|
Note this only appears if an Execute message's row-count limit
|
|
|
|
was reached.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-Query">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Query (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('Q')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a simple query.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-06-23 01:27:48 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The query string itself.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-ReadyForQuery">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
ReadyForQuery (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('Z')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message type. ReadyForQuery is sent
|
|
|
|
whenever the backend is ready for a new query cycle.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(5)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Current backend transaction status indicator.
|
2017-10-09 03:44:17 +02:00
|
|
|
Possible values are '<literal>I</literal>' if idle (not in
|
|
|
|
a transaction block); '<literal>T</literal>' if in a transaction
|
|
|
|
block; or '<literal>E</literal>' if in a failed transaction
|
2005-06-22 17:19:43 +02:00
|
|
|
block (queries will be rejected until block is ended).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-RowDescription">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
RowDescription (B)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-09-21 22:31:49 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('T')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a row description.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2001-09-21 22:31:49 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
Specifies the number of fields in a row (can be zero).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
Then, for each field, there is the following:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The field name.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
If the field can be identified as a column of a specific
|
2005-06-22 17:19:43 +02:00
|
|
|
table, the object ID of the table; otherwise zero.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
If the field can be identified as a column of a specific
|
2005-06-22 17:19:43 +02:00
|
|
|
table, the attribute number of the column; otherwise zero.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-11-01 02:56:29 +01:00
|
|
|
The object ID of the field's data type.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The data type size (see <varname>pg_type.typlen</varname>).
|
2005-06-22 17:19:43 +02:00
|
|
|
Note that negative values denote variable-width types.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The type modifier (see <varname>pg_attribute.atttypmod</varname>).
|
2005-06-22 17:19:43 +02:00
|
|
|
The meaning of the modifier is type-specific.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int16
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-05-07 23:46:15 +02:00
|
|
|
The format code being used for the field. Currently will
|
2005-06-22 17:19:43 +02:00
|
|
|
be zero (text) or one (binary). In a RowDescription
|
|
|
|
returned from the statement variant of Describe, the
|
|
|
|
format code is not yet known and will always be zero.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-SASLInitialResponse">
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
<term>
|
2017-08-23 18:01:43 +02:00
|
|
|
SASLInitialResponse (F)
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('p')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an initial SASL response. Note that
|
|
|
|
this is also used for GSSAPI, SSPI and password response messages.
|
|
|
|
The exact message type is deduced from the context.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Name of the SASL authentication mechanism that the client
|
|
|
|
selected.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of SASL mechanism specific "Initial Client Response" that
|
|
|
|
follows, or -1 if there is no Initial Response.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
SASL mechanism specific "Initial Response".
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-SASLResponse">
|
Improve the SASL authentication protocol.
This contains some protocol changes to SASL authentiation (which is new
in v10):
* For future-proofing, in the AuthenticationSASL message that begins SASL
authentication, provide a list of SASL mechanisms that the server
supports, for the client to choose from. Currently, it's always just
SCRAM-SHA-256.
* Add a separate authentication message type for the final server->client
SASL message, which the client doesn't need to respond to. This makes
it unambiguous whether the client is supposed to send a response or not.
The SASL mechanism should know that anyway, but better to be explicit.
Also, in the server, support clients that don't send an Initial Client
response in the first SASLInitialResponse message. The server is supposed
to first send an empty request in that case, to which the client will
respond with the data that usually comes in the Initial Client Response.
libpq uses the Initial Client Response field and doesn't need this, and I
would assume any other sensible implementation to use Initial Client
Response, too, but let's follow the SASL spec.
Improve the documentation on SASL authentication in protocol. Add a
section describing the SASL message flow, and some details on our
SCRAM-SHA-256 implementation.
Document the different kinds of PasswordMessages that the frontend sends
in different phases of SASL authentication, as well as GSS/SSPI
authentication as separate message formats. Even though they're all 'p'
messages, and the exact format depends on the context, describing them as
separate message formats makes the documentation more clear.
Reviewed by Michael Paquier and Álvaro Hernández Tortosa.
Discussion: https://www.postgresql.org/message-id/CAB7nPqS-aFg0iM3AQOJwKDv_0WkAedRjs1W2X8EixSz+sKBXCQ@mail.gmail.com
2017-04-13 18:34:16 +02:00
|
|
|
<term>
|
|
|
|
SASLResponse (F)
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('p')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a SASL response. Note that
|
|
|
|
this is also used for GSSAPI, SSPI and password response messages.
|
|
|
|
The exact message type can be deduced from the context.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of message contents in bytes, including self.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
SASL mechanism specific message data.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-SSLRequest">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
SSLRequest (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-07-09 05:29:11 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(8)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(80877103)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
The <acronym>SSL</acronym> request code. The value is chosen to contain
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>1234</literal> in the most significant 16 bits, and <literal>5679</literal> in the
|
2016-05-04 03:06:25 +02:00
|
|
|
least significant 16 bits. (To avoid confusion, this code
|
2003-04-16 00:51:18 +02:00
|
|
|
must not be the same as any protocol version number.)
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-07-09 05:29:11 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-StartupMessage">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
StartupMessage (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-05-07 23:46:15 +02:00
|
|
|
Int32(196608)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The protocol version number. The most significant 16 bits are
|
2003-05-07 23:46:15 +02:00
|
|
|
the major version number (3 for the protocol described here).
|
2005-06-22 17:19:43 +02:00
|
|
|
The least significant 16 bits are the minor version number
|
|
|
|
(0 for the protocol described here).
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
The protocol version number is followed by one or more pairs of
|
2005-06-22 17:19:43 +02:00
|
|
|
parameter name and value strings. A zero byte is required as a
|
|
|
|
terminator after the last name/value pair.
|
|
|
|
Parameters can appear in any
|
2017-10-09 03:44:17 +02:00
|
|
|
order. <literal>user</literal> is required, others are optional.
|
2005-06-22 17:19:43 +02:00
|
|
|
Each parameter is specified as:
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The parameter name. Currently recognized names are:
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>user</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The database user name to connect as. Required;
|
2005-06-22 17:19:43 +02:00
|
|
|
there is no default.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>database</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The database to connect to. Defaults to the user name.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>options</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Command-line arguments for the backend. (This is
|
2005-06-22 17:19:43 +02:00
|
|
|
deprecated in favor of setting individual run-time
|
2015-06-29 18:42:52 +02:00
|
|
|
parameters.) Spaces within this string are
|
|
|
|
considered to separate arguments, unless escaped with
|
2017-10-09 03:44:17 +02:00
|
|
|
a backslash (<literal>\</literal>); write <literal>\\</literal> to
|
2015-06-29 18:42:52 +02:00
|
|
|
represent a literal backslash.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2017-04-10 13:08:20 +02:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>replication</literal>
|
2017-04-10 13:08:20 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-04-10 13:13:15 +02:00
|
|
|
Used to connect in streaming replication mode, where
|
|
|
|
a small set of replication commands can be issued
|
|
|
|
instead of SQL statements. Value can be
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>true</literal>, <literal>false</literal>, or
|
|
|
|
<literal>database</literal>, and the default is
|
|
|
|
<literal>false</literal>. See
|
2017-11-23 15:39:47 +01:00
|
|
|
<xref linkend="protocol-replication"/> for details.
|
2017-04-10 13:08:20 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2005-01-23 01:30:59 +01:00
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2018-03-21 15:08:43 +01:00
|
|
|
In addition to the above, other parameters may be listed.
|
2017-11-21 19:56:24 +01:00
|
|
|
Parameter names beginning with <literal>_pq_.</literal> are
|
|
|
|
reserved for use as protocol extensions, while others are
|
|
|
|
treated as run-time parameters to be set at backend start
|
|
|
|
time. Such settings will be applied during backend start
|
|
|
|
(after parsing the command-line arguments if any) and will
|
|
|
|
act as session defaults.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
String
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
The parameter value.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-Sync">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Sync (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('S')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a Sync command.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2001-11-22 02:22:10 +01:00
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-message-formats-Terminate">
|
2005-01-23 01:30:59 +01:00
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Terminate (F)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Byte1('X')
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Identifies the message as a termination.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2003-04-16 00:51:18 +02:00
|
|
|
Int32(4)
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Length of message contents in bytes, including self.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
</sect1>
|
|
|
|
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<sect1 id="protocol-error-fields">
|
|
|
|
<title>Error and Notice Message Fields</title>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
|
|
|
<para>
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
This section describes the fields that can appear in ErrorResponse and
|
2003-04-16 00:51:18 +02:00
|
|
|
NoticeResponse messages. Each field type has a single-byte identification
|
2003-04-22 02:08:07 +02:00
|
|
|
token. Note that any given field type should appear at most once per
|
|
|
|
message.
|
2003-04-16 00:51:18 +02:00
|
|
|
</para>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<variablelist>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>S</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Severity: the field contents are
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>ERROR</literal>, <literal>FATAL</literal>, or
|
|
|
|
<literal>PANIC</literal> (in an error message), or
|
|
|
|
<literal>WARNING</literal>, <literal>NOTICE</literal>, <literal>DEBUG</literal>,
|
|
|
|
<literal>INFO</literal>, or <literal>LOG</literal> (in a notice message),
|
2005-06-22 17:19:43 +02:00
|
|
|
or a localized translation of one of these. Always present.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2016-08-26 22:20:17 +02:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>V</literal>
|
2016-08-26 22:20:17 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Severity: the field contents are
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>ERROR</literal>, <literal>FATAL</literal>, or
|
|
|
|
<literal>PANIC</literal> (in an error message), or
|
|
|
|
<literal>WARNING</literal>, <literal>NOTICE</literal>, <literal>DEBUG</literal>,
|
|
|
|
<literal>INFO</literal>, or <literal>LOG</literal> (in a notice message).
|
|
|
|
This is identical to the <literal>S</literal> field except
|
2016-08-26 22:20:17 +02:00
|
|
|
that the contents are never localized. This is present only in
|
2017-10-09 03:44:17 +02:00
|
|
|
messages generated by <productname>PostgreSQL</productname> versions 9.6
|
2016-08-26 22:20:17 +02:00
|
|
|
and later.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>C</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-10-17 20:57:01 +02:00
|
|
|
Code: the SQLSTATE code for the error (see <xref
|
2017-11-23 15:39:47 +01:00
|
|
|
linkend="errcodes-appendix"/>). Not localizable. Always present.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>M</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Message: the primary human-readable error message.
|
|
|
|
This should be accurate but terse (typically one line).
|
|
|
|
Always present.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>D</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Detail: an optional secondary error message carrying more
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
detail about the problem. Might run to multiple lines.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>H</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Hint: an optional suggestion what to do about the problem.
|
2005-06-22 17:19:43 +02:00
|
|
|
This is intended to differ from Detail in that it offers advice
|
|
|
|
(potentially inappropriate) rather than hard facts.
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
Might run to multiple lines.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>P</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Position: the field value is a decimal ASCII integer, indicating
|
2005-06-22 17:19:43 +02:00
|
|
|
an error cursor position as an index into the original query string.
|
|
|
|
The first character has index 1, and positions are measured in
|
|
|
|
characters not bytes.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>p</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
Internal position: this is defined the same as the <literal>P</literal>
|
2004-03-21 23:29:11 +01:00
|
|
|
field, but it is used when the cursor position refers to an internally
|
|
|
|
generated command rather than the one submitted by the client.
|
2017-10-09 03:44:17 +02:00
|
|
|
The <literal>q</literal> field will always appear when this field appears.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2004-03-21 23:29:11 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>q</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2004-03-21 23:29:11 +01:00
|
|
|
Internal query: the text of a failed internally-generated command.
|
2021-06-11 03:38:04 +02:00
|
|
|
This could be, for example, an SQL query issued by a PL/pgSQL function.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2004-03-21 23:29:11 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>W</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Where: an indication of the context in which the error occurred.
|
2004-03-21 23:29:11 +01:00
|
|
|
Presently this includes a call stack traceback of active
|
|
|
|
procedural language functions and internally-generated queries.
|
|
|
|
The trace is one entry per line, most recent first.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>s</literal>
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Schema name: if the error was associated with a specific database
|
|
|
|
object, the name of the schema containing that object, if any.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>t</literal>
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Table name: if the error was associated with a specific table, the
|
2013-07-03 13:29:23 +02:00
|
|
|
name of the table. (Refer to the schema name field for the name of
|
|
|
|
the table's schema.)
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>c</literal>
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Column name: if the error was associated with a specific table column,
|
2013-07-03 13:29:23 +02:00
|
|
|
the name of the column. (Refer to the schema and table name fields to
|
|
|
|
identify the table.)
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>d</literal>
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2013-05-21 03:13:13 +02:00
|
|
|
Data type name: if the error was associated with a specific data type,
|
2013-07-03 13:29:23 +02:00
|
|
|
the name of the data type. (Refer to the schema name field for the
|
|
|
|
name of the data type's schema.)
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>n</literal>
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Constraint name: if the error was associated with a specific
|
2013-07-03 13:29:23 +02:00
|
|
|
constraint, the name of the constraint. Refer to fields listed above
|
|
|
|
for the associated table or domain. (For this purpose, indexes are
|
|
|
|
treated as constraints, even if they weren't created with constraint
|
|
|
|
syntax.)
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>F</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
File: the file name of the source-code location where the error
|
2005-06-22 17:19:43 +02:00
|
|
|
was reported.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>L</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Line: the line number of the source-code location where the error
|
2005-06-22 17:19:43 +02:00
|
|
|
was reported.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-10-09 03:44:17 +02:00
|
|
|
<literal>R</literal>
|
2005-01-23 01:30:59 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2003-04-16 00:51:18 +02:00
|
|
|
Routine: the name of the source-code routine reporting the error.
|
2005-01-23 01:30:59 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</variablelist>
|
1998-03-01 09:16:16 +01:00
|
|
|
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
<note>
|
|
|
|
<para>
|
2013-05-21 03:13:13 +02:00
|
|
|
The fields for schema name, table name, column name, data type name, and
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
constraint name are supplied only for a limited number of error types;
|
2017-11-23 15:39:47 +01:00
|
|
|
see <xref linkend="errcodes-appendix"/>. Frontends should not assume that
|
2013-07-03 13:29:23 +02:00
|
|
|
the presence of any of these fields guarantees the presence of another
|
|
|
|
field. Core error sources observe the interrelationships noted above, but
|
|
|
|
user-defined functions may use these fields in other ways. In the same
|
|
|
|
vein, clients should not assume that these fields denote contemporary
|
|
|
|
objects in the current database.
|
Provide database object names as separate fields in error messages.
This patch addresses the problem that applications currently have to
extract object names from possibly-localized textual error messages,
if they want to know for example which index caused a UNIQUE_VIOLATION
failure. It adds new error message fields to the wire protocol, which
can carry the name of a table, table column, data type, or constraint
associated with the error. (Since the protocol spec has always instructed
clients to ignore unrecognized field types, this should not create any
compatibility problem.)
Support for providing these new fields has been added to just a limited set
of error reports (mainly, those in the "integrity constraint violation"
SQLSTATE class), but we will doubtless add them to more calls in future.
Pavel Stehule, reviewed and extensively revised by Peter Geoghegan, with
additional hacking by Tom Lane.
2013-01-29 23:06:26 +01:00
|
|
|
</para>
|
|
|
|
</note>
|
|
|
|
|
2003-04-16 00:51:18 +02:00
|
|
|
<para>
|
|
|
|
The client is responsible for formatting displayed information to meet its
|
|
|
|
needs; in particular it should break long lines as needed. Newline characters
|
|
|
|
appearing in the error message fields should be treated as paragraph breaks,
|
|
|
|
not line breaks.
|
|
|
|
</para>
|
|
|
|
|
1998-12-29 03:24:47 +01:00
|
|
|
</sect1>
|
2003-04-16 00:51:18 +02:00
|
|
|
|
2017-01-19 18:00:00 +01:00
|
|
|
<sect1 id="protocol-logicalrep-message-formats">
|
|
|
|
<title>Logical Replication Message Formats</title>
|
|
|
|
|
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
This section describes the detailed format of each logical replication
|
|
|
|
message. These messages are either returned by the replication slot SQL
|
2022-01-25 01:40:04 +01:00
|
|
|
interface or are sent by a walsender. In the case of a walsender, they are
|
2021-08-02 17:32:17 +02:00
|
|
|
encapsulated inside replication protocol WAL messages as described in
|
|
|
|
<xref linkend="protocol-replication"/>, and generally obey the same message
|
|
|
|
flow as physical replication.
|
2017-01-19 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Begin">
|
2017-01-19 18:00:00 +01:00
|
|
|
<term>
|
|
|
|
Begin
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('B')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a begin message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (XLogRecPtr)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The final LSN of the transaction.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (TimestampTz)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Commit timestamp of the transaction. The value is in number
|
|
|
|
of microseconds since PostgreSQL epoch (2000-01-01).
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Xid of the transaction.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
2021-04-06 05:10:47 +02:00
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Message">
|
2021-04-06 05:10:47 +02:00
|
|
|
<term>
|
|
|
|
Message
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('M')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a logical decoding message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2021-04-06 05:10:47 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2021-04-12 05:57:10 +02:00
|
|
|
Xid of the transaction (only present for streamed transactions).
|
|
|
|
This field is available since protocol version 2.
|
2021-04-06 05:10:47 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int8
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Flags; Either 0 for no flags or 1 if the logical decoding
|
|
|
|
message is transactional.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (XLogRecPtr)
|
2021-04-06 05:10:47 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The LSN of the logical decoding message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The prefix of the logical decoding message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2021-06-24 08:21:58 +02:00
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of the content.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2021-04-06 05:10:47 +02:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte<replaceable>n</replaceable>
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The content of the logical decoding message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
2017-01-19 18:00:00 +01:00
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Commit">
|
2017-01-19 18:00:00 +01:00
|
|
|
<term>
|
|
|
|
Commit
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('C')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a commit message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
2017-08-14 19:42:03 +02:00
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int8(0)
|
2017-08-14 19:42:03 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
Flags; currently unused.
|
2017-08-14 19:42:03 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
2017-01-19 18:00:00 +01:00
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (XLogRecPtr)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The LSN of the commit.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (XLogRecPtr)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The end LSN of the transaction.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (TimestampTz)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Commit timestamp of the transaction. The value is in number
|
|
|
|
of microseconds since PostgreSQL epoch (2000-01-01).
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Origin">
|
2017-01-19 18:00:00 +01:00
|
|
|
<term>
|
|
|
|
Origin
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('O')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an origin message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (XLogRecPtr)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The LSN of the commit on the origin server.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Name of the origin.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
Note that there can be multiple Origin messages inside a single transaction.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Relation">
|
2017-01-19 18:00:00 +01:00
|
|
|
<term>
|
|
|
|
Relation
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('R')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a relation message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2021-04-12 05:57:10 +02:00
|
|
|
<para>
|
|
|
|
Xid of the transaction (only present for streamed transactions).
|
|
|
|
This field is available since protocol version 2.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (Oid)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2017-01-19 18:00:00 +01:00
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
OID of the relation.
|
2017-01-19 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Namespace (empty string for <literal>pg_catalog</literal>).
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Relation name.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int8
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Replica identity setting for the relation (same as
|
|
|
|
<structfield>relreplident</structfield> in <structname>pg_class</structname>).
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int16
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Number of columns.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
Allow specifying column lists for logical replication
This allows specifying an optional column list when adding a table to
logical replication. The column list may be specified after the table
name, enclosed in parentheses. Columns not included in this list are not
sent to the subscriber, allowing the schema on the subscriber to be a
subset of the publisher schema.
For UPDATE/DELETE publications, the column list needs to cover all
REPLICA IDENTITY columns. For INSERT publications, the column list is
arbitrary and may omit some REPLICA IDENTITY columns. Furthermore, if
the table uses REPLICA IDENTITY FULL, column list is not allowed.
The column list can contain only simple column references. Complex
expressions, function calls etc. are not allowed. This restriction could
be relaxed in the future.
During the initial table synchronization, only columns included in the
column list are copied to the subscriber. If the subscription has
several publications, containing the same table with different column
lists, columns specified in any of the lists will be copied.
This means all columns are replicated if the table has no column list
at all (which is treated as column list with all columns), or when of
the publications is defined as FOR ALL TABLES (possibly IN SCHEMA that
matches the schema of the table).
For partitioned tables, publish_via_partition_root determines whether
the column list for the root or the leaf relation will be used. If the
parameter is 'false' (the default), the list defined for the leaf
relation is used. Otherwise, the column list for the root partition
will be used.
Psql commands \dRp+ and \d <table-name> now display any column lists.
Author: Tomas Vondra, Alvaro Herrera, Rahila Syed
Reviewed-by: Peter Eisentraut, Alvaro Herrera, Vignesh C, Ibrar Ahmed,
Amit Kapila, Hou zj, Peter Smith, Wang wei, Tang, Shi yu
Discussion: https://postgr.es/m/CAH2L28vddB_NFdRVpuyRBJEBWjz4BSyTB=_ektNRH8NJ1jf95g@mail.gmail.com
2022-03-26 00:45:21 +01:00
|
|
|
Next, the following message part appears for each column included in
|
|
|
|
the publication (except generated columns):
|
2017-01-19 18:00:00 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int8
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Flags for the column. Currently can be either 0 for no flags
|
|
|
|
or 1 which marks the column as part of the key.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Name of the column.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2017-08-09 01:18:16 +02:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (Oid)
|
2017-08-09 01:18:16 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
OID of the column's data type.
|
2017-08-09 01:18:16 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Type modifier of the column (<structfield>atttypmod</structfield>).
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2017-01-19 18:00:00 +01:00
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Type">
|
2017-08-15 21:36:18 +02:00
|
|
|
<term>
|
|
|
|
Type
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('Y')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a type message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2017-08-15 21:36:18 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2021-04-12 05:57:10 +02:00
|
|
|
<para>
|
|
|
|
Xid of the transaction (only present for streamed transactions).
|
|
|
|
This field is available since protocol version 2.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (Oid)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2017-08-15 21:36:18 +02:00
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
OID of the data type.
|
2017-08-15 21:36:18 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Namespace (empty string for <literal>pg_catalog</literal>).
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
String
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Name of the data type.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Insert">
|
2017-01-19 18:00:00 +01:00
|
|
|
<term>
|
|
|
|
Insert
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('I')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an insert message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2021-04-12 05:57:10 +02:00
|
|
|
<para>
|
|
|
|
Xid of the transaction (only present for streamed transactions).
|
|
|
|
This field is available since protocol version 2.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (Oid)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2017-01-19 18:00:00 +01:00
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
OID of the relation corresponding to the ID in the relation
|
2017-01-19 18:00:00 +01:00
|
|
|
message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('N')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the following TupleData message as a new tuple.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
TupleData
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
TupleData message part representing the contents of new tuple.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Update">
|
2017-01-19 18:00:00 +01:00
|
|
|
<term>
|
|
|
|
Update
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('U')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as an update message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2021-04-12 05:57:10 +02:00
|
|
|
<para>
|
|
|
|
Xid of the transaction (only present for streamed transactions).
|
|
|
|
This field is available since protocol version 2.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (Oid)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2017-01-19 18:00:00 +01:00
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
OID of the relation corresponding to the ID in the relation
|
2017-01-19 18:00:00 +01:00
|
|
|
message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('K')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the following TupleData submessage as a key.
|
|
|
|
This field is optional and is only present if
|
|
|
|
the update changed data in any of the column(s) that are
|
|
|
|
part of the REPLICA IDENTITY index.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('O')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the following TupleData submessage as an old tuple.
|
|
|
|
This field is optional and is only present if table in which
|
|
|
|
the update happened has REPLICA IDENTITY set to FULL.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
TupleData
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
TupleData message part representing the contents of the old tuple
|
|
|
|
or primary key. Only present if the previous 'O' or 'K' part
|
|
|
|
is present.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('N')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the following TupleData message as a new tuple.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
TupleData
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
TupleData message part representing the contents of a new tuple.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The Update message may contain either a 'K' message part or an 'O' message part
|
|
|
|
or neither of them, but never both of them.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Delete">
|
2017-01-19 18:00:00 +01:00
|
|
|
<term>
|
|
|
|
Delete
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('D')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a delete message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2021-04-12 05:57:10 +02:00
|
|
|
<para>
|
|
|
|
Xid of the transaction (only present for streamed transactions).
|
|
|
|
This field is available since protocol version 2.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (Oid)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2017-01-19 18:00:00 +01:00
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
OID of the relation corresponding to the ID in the relation
|
2017-01-19 18:00:00 +01:00
|
|
|
message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('K')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the following TupleData submessage as a key.
|
|
|
|
This field is present if the table in which the delete has
|
|
|
|
happened uses an index as REPLICA IDENTITY.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('O')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2021-02-24 08:13:17 +01:00
|
|
|
Identifies the following TupleData message as an old tuple.
|
|
|
|
This field is present if the table in which the delete
|
2017-01-19 18:00:00 +01:00
|
|
|
happened has REPLICA IDENTITY set to FULL.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
TupleData
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
TupleData message part representing the contents of the old tuple
|
|
|
|
or primary key, depending on the previous field.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The Delete message may contain either a 'K' message part or an 'O' message part,
|
|
|
|
but never both of them.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Truncate">
|
2018-04-07 17:24:53 +02:00
|
|
|
<term>
|
|
|
|
Truncate
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('T')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a truncate message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2018-04-07 17:24:53 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
2021-04-12 05:57:10 +02:00
|
|
|
<para>
|
|
|
|
Xid of the transaction (only present for streamed transactions).
|
|
|
|
This field is available since protocol version 2.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
2018-04-07 17:24:53 +02:00
|
|
|
<para>
|
|
|
|
Number of relations
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int8
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Option bits for <command>TRUNCATE</command>:
|
|
|
|
1 for <literal>CASCADE</literal>, 2 for <literal>RESTART IDENTITY</literal>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (Oid)
|
2018-04-07 17:24:53 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
OID of the relation corresponding to the ID in the relation
|
2018-04-07 17:24:53 +02:00
|
|
|
message. This field is repeated for each relation.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2017-01-19 18:00:00 +01:00
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
|
2021-08-19 05:33:11 +02:00
|
|
|
The following messages (Stream Start, Stream Stop, Stream Commit, and
|
2021-04-12 05:57:10 +02:00
|
|
|
Stream Abort) are available since protocol version 2.
|
|
|
|
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Stream-Start">
|
2021-04-12 05:57:10 +02:00
|
|
|
<term>
|
|
|
|
Stream Start
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('S')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a stream start message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Xid of the transaction.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int8
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
A value of 1 indicates this is the first stream segment for
|
|
|
|
this XID, 0 for any other stream segment.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Stream-Stop">
|
2021-04-12 05:57:10 +02:00
|
|
|
<term>
|
|
|
|
Stream Stop
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('E')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a stream stop message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Stream-Commit">
|
2021-04-12 05:57:10 +02:00
|
|
|
<term>
|
|
|
|
Stream Commit
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('c')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a stream commit message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Xid of the transaction.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int8(0)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2021-08-02 17:32:17 +02:00
|
|
|
Flags; currently unused.
|
2021-04-12 05:57:10 +02:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (XLogRecPtr)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The LSN of the commit.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (XLogRecPtr)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
The end LSN of the transaction.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int64 (TimestampTz)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Commit timestamp of the transaction. The value is in number
|
|
|
|
of microseconds since PostgreSQL epoch (2000-01-01).
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Stream-Abort">
|
2021-04-12 05:57:10 +02:00
|
|
|
<term>
|
|
|
|
Stream Abort
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('A')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the message as a stream abort message.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Xid of the transaction.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2021-08-02 17:32:17 +02:00
|
|
|
Int32 (TransactionId)
|
2021-04-12 05:57:10 +02:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Xid of the subtransaction (will be same as xid of the transaction for top-level
|
|
|
|
transactions).
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<para>
|
2021-08-04 04:17:06 +02:00
|
|
|
The following messages (Begin Prepare, Prepare, Commit Prepared, Rollback Prepared, Stream Prepare)
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
are available since protocol version 3.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Begin-Prepare">
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
|
|
|
|
<term>Begin Prepare</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte1('b')</term>
|
|
|
|
<listitem><para>
|
|
|
|
Identifies the message as the beginning of a two-phase transaction message.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
The LSN of the prepare.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
The end LSN of the prepared transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (TimestampTz)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
Prepare timestamp of the transaction. The value is in number
|
|
|
|
of microseconds since PostgreSQL epoch (2000-01-01).
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int32 (TransactionId)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
Xid of the transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>String</term>
|
|
|
|
<listitem><para>
|
|
|
|
The user defined GID of the two-phase transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Prepare">
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
|
|
|
|
<term>Prepare</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte1('P')</term>
|
|
|
|
<listitem><para>
|
|
|
|
Identifies the message as a two-phase prepared transaction message.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int8(0)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
2021-08-02 17:32:17 +02:00
|
|
|
Flags; currently unused.
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
The LSN of the prepare.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
The end LSN of the prepared transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (TimestampTz)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
Prepare timestamp of the transaction. The value is in number
|
|
|
|
of microseconds since PostgreSQL epoch (2000-01-01).
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int32 (TransactionId)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
Xid of the transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>String</term>
|
|
|
|
<listitem><para>
|
|
|
|
The user defined GID of the two-phase transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Commit-Prepared">
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
|
|
|
|
<term>Commit Prepared</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte1('K')</term>
|
|
|
|
<listitem><para>
|
|
|
|
Identifies the message as the commit of a two-phase transaction message.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int8(0)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
2021-08-02 17:32:17 +02:00
|
|
|
Flags; currently unused.
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
The LSN of the commit prepared.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
The end LSN of the commit prepared transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (TimestampTz)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
Commit timestamp of the transaction. The value is in number
|
|
|
|
of microseconds since PostgreSQL epoch (2000-01-01).
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int32 (TransactionId)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
Xid of the transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>String</term>
|
|
|
|
<listitem><para>
|
|
|
|
The user defined GID of the two-phase transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Rollback-Prepared">
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
|
|
|
|
<term>Rollback Prepared</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte1('r')</term>
|
|
|
|
<listitem><para>
|
|
|
|
Identifies the message as the rollback of a two-phase transaction message.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int8(0)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
2021-08-02 17:32:17 +02:00
|
|
|
Flags; currently unused.
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
The end LSN of the prepared transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
The end LSN of the rollback prepared transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (TimestampTz)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
Prepare timestamp of the transaction. The value is in number
|
|
|
|
of microseconds since PostgreSQL epoch (2000-01-01).
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int64 (TimestampTz)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
Rollback timestamp of the transaction. The value is in number
|
|
|
|
of microseconds since PostgreSQL epoch (2000-01-01).
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
2021-08-02 17:32:17 +02:00
|
|
|
<term>
|
|
|
|
Int32 (TransactionId)
|
|
|
|
</term>
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<listitem><para>
|
|
|
|
Xid of the transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2021-08-04 04:17:06 +02:00
|
|
|
<varlistentry>
|
|
|
|
<term>String</term>
|
|
|
|
<listitem><para>
|
|
|
|
The user defined GID of the two-phase transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-Stream-Prepare">
|
2021-08-04 04:17:06 +02:00
|
|
|
|
|
|
|
<term>Stream Prepare</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>Byte1('p')</term>
|
|
|
|
<listitem><para>
|
|
|
|
Identifies the message as a two-phase stream prepare message.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int8(0)
|
|
|
|
</term>
|
|
|
|
<listitem><para>
|
|
|
|
Flags; currently unused.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
|
|
|
<listitem><para>
|
|
|
|
The LSN of the prepare.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int64 (XLogRecPtr)
|
|
|
|
</term>
|
|
|
|
<listitem><para>
|
|
|
|
The end LSN of the prepare transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int64 (TimestampTz)
|
|
|
|
</term>
|
|
|
|
<listitem><para>
|
|
|
|
Prepare timestamp of the transaction. The value is in number
|
|
|
|
of microseconds since PostgreSQL epoch (2000-01-01).
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32 (TransactionId)
|
|
|
|
</term>
|
|
|
|
<listitem><para>
|
|
|
|
Xid of the transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
Add support for prepared transactions to built-in logical replication.
To add support for streaming transactions at prepare time into the
built-in logical replication, we need to do the following things:
* Modify the output plugin (pgoutput) to implement the new two-phase API
callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle two-phase
transactions by replaying them on prepare.
* Add a new SUBSCRIPTION option "two_phase" to allow users to enable
two-phase transactions. We enable the two_phase once the initial data sync
is over.
We however must explicitly disable replication of two-phase transactions
during replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover, we don't have a replication connection open so we don't know
where to send the data anyway.
The streaming option is not allowed with this new two_phase option. This
can be done as a separate patch.
We don't allow to toggle two_phase option of a subscription because it can
lead to an inconsistent replica. For the same reason, we don't allow to
refresh the publication once the two_phase is enabled for a subscription
unless copy_data option is false.
Author: Peter Smith, Ajin Cherian and Amit Kapila based on previous work by Nikhil Sontakke and Stas Kelvich
Reviewed-by: Amit Kapila, Sawada Masahiko, Vignesh C, Dilip Kumar, Takamichi Osumi, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAA4eK1+opiV4aFTmWWUF9h_32=HfPOW9vZASHarT0UA5oBrtGw@mail.gmail.com
2021-07-14 04:03:50 +02:00
|
|
|
<varlistentry>
|
|
|
|
<term>String</term>
|
|
|
|
<listitem><para>
|
|
|
|
The user defined GID of the two-phase transaction.
|
|
|
|
</para></listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
|
2021-04-12 05:57:10 +02:00
|
|
|
<para>
|
|
|
|
|
2017-08-09 01:18:16 +02:00
|
|
|
The following message parts are shared by the above messages.
|
2017-01-19 18:00:00 +01:00
|
|
|
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
|
2022-03-02 10:30:41 +01:00
|
|
|
<varlistentry id="protocol-logicalrep-message-formats-TupleData">
|
2017-01-19 18:00:00 +01:00
|
|
|
<term>
|
|
|
|
TupleData
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int16
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Number of columns.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
2019-03-30 08:13:09 +01:00
|
|
|
Next, one of the following submessages appears for each column (except generated columns):
|
2017-01-19 18:00:00 +01:00
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('n')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-03-14 17:57:10 +01:00
|
|
|
Identifies the data as NULL value.
|
2017-01-19 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
Or
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('u')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-03-14 17:57:10 +01:00
|
|
|
Identifies unchanged TOASTed value (the actual value is not
|
2017-01-19 18:00:00 +01:00
|
|
|
sent).
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
</variablelist>
|
|
|
|
Or
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('t')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2017-03-14 17:57:10 +01:00
|
|
|
Identifies the data as text formatted value.
|
2017-01-19 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2021-06-24 08:21:58 +02:00
|
|
|
</variablelist>
|
|
|
|
Or
|
|
|
|
<variablelist>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Byte1('b')
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Identifies the data as binary formatted value.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
2017-01-19 18:00:00 +01:00
|
|
|
<varlistentry>
|
|
|
|
<term>
|
|
|
|
Int32
|
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
|
|
|
Length of the column value.
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
<varlistentry>
|
|
|
|
<term>
|
2017-04-05 16:44:23 +02:00
|
|
|
Byte<replaceable>n</replaceable>
|
2017-01-19 18:00:00 +01:00
|
|
|
</term>
|
|
|
|
<listitem>
|
|
|
|
<para>
|
2021-06-24 08:21:58 +02:00
|
|
|
The value of the column, either in binary or in text format.
|
|
|
|
(As specified in the preceding format byte).
|
2017-04-05 16:44:23 +02:00
|
|
|
<replaceable>n</replaceable> is the above length.
|
|
|
|
|
2017-01-19 18:00:00 +01:00
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
</para>
|
|
|
|
</listitem>
|
|
|
|
</varlistentry>
|
|
|
|
|
|
|
|
</variablelist>
|
|
|
|
|
|
|
|
</sect1>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
<sect1 id="protocol-changes">
|
|
|
|
<title>Summary of Changes since Protocol 2.0</title>
|
2003-04-19 02:02:30 +02:00
|
|
|
|
|
|
|
<para>
|
|
|
|
This section provides a quick checklist of changes, for the benefit of
|
|
|
|
developers trying to update existing client libraries to protocol 3.0.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The initial startup packet uses a flexible list-of-strings format
|
|
|
|
instead of a fixed format. Notice that session default values for run-time
|
|
|
|
parameters can now be specified directly in the startup packet. (Actually,
|
2017-10-09 03:44:17 +02:00
|
|
|
you could do that before using the <literal>options</literal> field, but given the
|
|
|
|
limited width of <literal>options</literal> and the lack of any way to quote
|
2003-04-19 02:02:30 +02:00
|
|
|
whitespace in the values, it wasn't a very safe technique.)
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
All messages now have a length count immediately following the message type
|
|
|
|
byte (except for startup packets, which have no type byte). Also note that
|
|
|
|
PasswordMessage now has a type byte.
|
|
|
|
</para>
|
|
|
|
|
2003-04-24 23:16:45 +02:00
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
ErrorResponse and NoticeResponse ('<literal>E</literal>' and '<literal>N</literal>')
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
messages now contain multiple fields, from which the client code can
|
2003-04-24 23:16:45 +02:00
|
|
|
assemble an error message of the desired level of verbosity. Note that
|
|
|
|
individual fields will typically not end with a newline, whereas the single
|
|
|
|
string sent in the older protocol always did.
|
|
|
|
</para>
|
|
|
|
|
2003-04-26 22:23:00 +02:00
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The ReadyForQuery ('<literal>Z</literal>') message includes a transaction status
|
2003-04-26 22:23:00 +02:00
|
|
|
indicator.
|
|
|
|
</para>
|
|
|
|
|
2003-05-07 23:46:15 +02:00
|
|
|
<para>
|
|
|
|
The distinction between BinaryRow and DataRow message types is gone; the
|
|
|
|
single DataRow message type serves for returning data in all formats.
|
|
|
|
Note that the layout of DataRow has changed to make it easier to parse.
|
|
|
|
Also, the representation of binary values has changed: it is no longer
|
|
|
|
directly tied to the server's internal representation.
|
|
|
|
</para>
|
|
|
|
|
2003-05-05 02:44:56 +02:00
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
There is a new <quote>extended query</quote> sub-protocol, which adds the frontend
|
2003-05-05 02:44:56 +02:00
|
|
|
message types Parse, Bind, Execute, Describe, Close, Flush, and Sync, and the
|
|
|
|
backend message types ParseComplete, BindComplete, PortalSuspended,
|
|
|
|
ParameterDescription, NoData, and CloseComplete. Existing clients do not
|
|
|
|
have to concern themselves with this sub-protocol, but making use of it
|
Update documentation on may/can/might:
Standard English uses "may", "can", and "might" in different ways:
may - permission, "You may borrow my rake."
can - ability, "I can lift that log."
might - possibility, "It might rain today."
Unfortunately, in conversational English, their use is often mixed, as
in, "You may use this variable to do X", when in fact, "can" is a better
choice. Similarly, "It may crash" is better stated, "It might crash".
Also update two error messages mentioned in the documenation to match.
2007-01-31 21:56:20 +01:00
|
|
|
might allow improvements in performance or functionality.
|
2003-05-05 02:44:56 +02:00
|
|
|
</para>
|
|
|
|
|
2003-04-19 02:02:30 +02:00
|
|
|
<para>
|
2003-12-14 01:10:32 +01:00
|
|
|
<command>COPY</command> data is now encapsulated into CopyData and CopyDone messages. There
|
|
|
|
is a well-defined way to recover from errors during <command>COPY</command>. The special
|
2017-10-09 03:44:17 +02:00
|
|
|
<quote><literal>\.</literal></quote> last line is not needed anymore, and is not sent
|
2003-12-14 01:10:32 +01:00
|
|
|
during <command>COPY OUT</command>.
|
|
|
|
(It is still recognized as a terminator during <command>COPY IN</command>, but its use is
|
|
|
|
deprecated and will eventually be removed.) Binary <command>COPY</command> is supported.
|
2003-05-07 23:46:15 +02:00
|
|
|
The CopyInResponse and CopyOutResponse messages include fields indicating
|
|
|
|
the number of columns and the format of each column.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
|
|
|
The layout of FunctionCall and FunctionCallResponse messages has changed.
|
|
|
|
FunctionCall can now support passing NULL arguments to functions. It also
|
|
|
|
can handle passing parameters and retrieving results in either text or
|
|
|
|
binary format. There is no longer any reason to consider FunctionCall a
|
|
|
|
potential security hole, since it does not offer direct access to internal
|
|
|
|
server data representations.
|
2003-04-22 02:08:07 +02:00
|
|
|
</para>
|
|
|
|
|
2003-04-25 21:45:10 +02:00
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The backend sends ParameterStatus ('<literal>S</literal>') messages during connection
|
2003-04-25 21:45:10 +02:00
|
|
|
startup for all parameters it considers interesting to the client library.
|
|
|
|
Subsequently, a ParameterStatus message is sent whenever the active value
|
|
|
|
changes for any of these parameters.
|
|
|
|
</para>
|
|
|
|
|
2003-04-26 22:23:00 +02:00
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The RowDescription ('<literal>T</literal>') message carries new table OID and column
|
2003-05-07 23:46:15 +02:00
|
|
|
number fields for each column of the described row. It also shows the format
|
|
|
|
code for each column.
|
2003-04-26 22:23:00 +02:00
|
|
|
</para>
|
|
|
|
|
2003-04-22 02:08:07 +02:00
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The CursorResponse ('<literal>P</literal>') message is no longer generated by
|
2003-04-22 02:08:07 +02:00
|
|
|
the backend.
|
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The NotificationResponse ('<literal>A</literal>') message has an additional string
|
|
|
|
field, which can carry a <quote>payload</quote> string passed
|
2003-12-14 01:10:32 +01:00
|
|
|
from the <command>NOTIFY</command> event sender.
|
2003-04-22 02:08:07 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
<para>
|
2017-10-09 03:44:17 +02:00
|
|
|
The EmptyQueryResponse ('<literal>I</literal>') message used to include an empty
|
2003-04-22 02:08:07 +02:00
|
|
|
string parameter; this has been removed.
|
2003-04-19 02:02:30 +02:00
|
|
|
</para>
|
|
|
|
|
|
|
|
</sect1>
|
|
|
|
|
2005-01-23 01:30:59 +01:00
|
|
|
</chapter>
|