The table column that stores this is of type oid, but is actually limited
to uint16 and has a different path for creating new values. Some of
the documentation already referred to it as an ID, so let's standardize
on that.
While at it, most format strings already use %u, so for consintency
change the remaining stragglers using %d.
Per suggestions from Tom Lane and Justin Pryzby
Discussion: https://www.postgresql.org/message-id/3437166.1659620465%40sss.pgh.pa.us
Backpatch to v15
The standard way to check for list emptiness is to compare the
List pointer to NIL; our list code goes out of its way to ensure
that that is the only representation of an empty list. (An
acceptable alternative is a plain boolean test for non-null
pointer, but explicit mention of NIL is usually preferable.)
Various places didn't get that memo and expressed the condition
with list_length(), which might not be so bad except that there
were such a variety of ways to check it exactly: equal to zero,
less than or equal to zero, less than one, yadda yadda. In the
name of code readability, let's standardize all those spellings
as "list == NIL" or "list != NIL". (There's probably some
microscopic efficiency gain too, though few of these look to be
at all performance-critical.)
A very small number of cases were left as-is because they seemed
more consistent with other adjacent list_length tests that way.
Peter Smith, with bikeshedding from a number of us
Discussion: https://postgr.es/m/CAHut+PtQYe+ENX5KrONMfugf0q6NHg4hR5dAhqEXEc2eefFeig@mail.gmail.com
Previously, we relied on HEAP2_NEW_CID records and XACT_INVALIDATION
records to know if the transaction has modified the catalog, and that
information is not serialized to snapshot. Therefore, after the restart,
if the logical decoding decodes only the commit record of the transaction
that has actually modified a catalog, we will miss adding its XID to the
snapshot. Thus, we will end up looking at catalogs with the wrong
snapshot.
To fix this problem, this change adds the list of transaction IDs and
sub-transaction IDs, that have modified catalogs and are running during
snapshot serialization, to the serialized snapshot. After restart or
otherwise, when we restore from such a serialized snapshot, the
corresponding list is restored in memory. Now, when decoding a COMMIT
record, we check both the list and the ReorderBuffer to see if the
transaction has modified catalogs.
Since this adds additional information to the serialized snapshot, we
cannot backpatch it. For back branches, we took another approach.
We remember the last-running-xacts list of the decoded RUNNING_XACTS
record after restoring the previously serialized snapshot. Then, we mark
the transaction as containing catalog changes if it's in the list of
initial running transactions and its commit record has
XACT_XINFO_HAS_INVALS. This doesn't require any file format changes but
the transaction will end up being added to the snapshot even if it has
only relcache invalidations. But that won't be a problem since we use
snapshot built during decoding only to read system catalogs.
This commit bumps SNAPBUILD_VERSION because of a change in SnapBuild.
Reported-by: Mike Oh
Author: Masahiko Sawada
Reviewed-by: Amit Kapila, Shi yu, Takamichi Osumi, Kyotaro Horiguchi, Bertrand Drouvot, Ahsan Hadi
Backpatch-through: 10
Discussion: https://postgr.es/m/81D0D8B0-E7C4-4999-B616-1E5004DBDCD2%40amazon.com
GetSubscriptionRelations() and GetSubscriptionNotReadyRelations() share
mostly the same code, which scans pg_subscription_rel and fetches all
the relations of a given subscription. The only difference is that the
second routine looks for all the relations not in a ready state. This
commit refactors the code to use a single routine, shaving a bit of
code.
Author: Vignesh C
Reviewed-By: Kyotaro Horiguchi, Amit Kapila, Michael Paquier, Peter
Smith
Discussion: https://postgr.es/m/CALDaNm0eW-9g4G_EzHebnFT5zZoasWCS_EzZQ5BgnLZny9S=pg@mail.gmail.com
This patch adds a new SUBSCRIPTION parameter "origin". It specifies
whether the subscription will request the publisher to only send changes
that don't have an origin or send changes regardless of origin. Setting it
to "none" means that the subscription will request the publisher to only
send changes that have no origin associated. Setting it to "any" means
that the publisher sends changes regardless of their origin. The default
is "any".
Usage:
CREATE SUBSCRIPTION sub1 CONNECTION 'dbname=postgres port=9999'
PUBLICATION pub1 WITH (origin = none);
This can be used to avoid loops (infinite replication of the same data)
among replication nodes.
This feature allows filtering only the replication data originating from
WAL but for initial sync (initial copy of table data) we don't have such a
facility as we can only distinguish the data based on origin from WAL. As
a follow-up patch, we are planning to forbid the initial sync if the
origin is specified as none and we notice that the publication tables were
also replicated from other publishers to avoid duplicate data or loops.
We forbid to allow creating origin with names 'none' and 'any' to avoid
confusion with the same name options.
Author: Vignesh C, Amit Kapila
Reviewed-By: Peter Smith, Amit Kapila, Dilip Kumar, Shi yu, Ashutosh Bapat, Hayato Kuroda
Discussion: https://postgr.es/m/CALDaNm0gwjY_4HFxvvty01BOT01q_fJLKQ3pWP9=9orqubhjcQ@mail.gmail.com
This replaces all MemSet() calls with struct initialization where that
is easily and obviously possible. (For example, some cases have to
worry about padding bits, so I left those.)
(The same could be done with appropriate memset() calls, but this
patch is part of an effort to phase out MemSet(), so it doesn't touch
memset() calls.)
Reviewed-by: Ranier Vilela <ranier.vf@gmail.com>
Reviewed-by: Alvaro Herrera <alvherre@alvh.no-ip.org>
Discussion: https://www.postgresql.org/message-id/9847b13c-b785-f4e2-75c3-12ec77a3b05c@enterprisedb.com
We have been using the term RelFileNode to refer to either (1) the
integer that is used to name the sequence of files for a certain relation
within the directory set aside for that tablespace/database combination;
or (2) that value plus the OIDs of the tablespace and database; or
occasionally (3) the whole series of files created for a relation
based on those values. Using the same name for more than one thing is
confusing.
Replace RelFileNode with RelFileNumber when we're talking about just the
single number, i.e. (1) from above, and with RelFileLocator when we're
talking about all the things that are needed to locate a relation's files
on disk, i.e. (2) from above. In the places where we refer to (3) as
a relfilenode, instead refer to "relation storage".
Since there is a ton of SQL code in the world that knows about
pg_class.relfilenode, don't change the name of that column, or of other
SQL-facing things that derive their name from it.
On the other hand, do adjust closely-related internal terminology. For
example, the structure member names dbNode and spcNode appear to be
derived from the fact that the structure itself was called RelFileNode,
so change those to dbOid and spcOid. Likewise, various variables with
names like rnode and relnode get renamed appropriately, according to
how they're being used in context.
Hopefully, this is clearer than before. It is also preparation for
future patches that intend to widen the relfilenumber fields from its
current width of 32 bits. Variables that store a relfilenumber are now
declared as type RelFileNumber rather than type Oid; right now, these
are the same, but that can now more easily be changed.
Dilip Kumar, per an idea from me. Reviewed also by Andres Freund.
I fixed some whitespace issues, changed a couple of words in a
comment, and made one other minor correction.
Discussion: http://postgr.es/m/CA+TgmoamOtXbVAQf9hWFzonUo6bhhjS6toZQd7HZ-pmojtAmag@mail.gmail.com
Discussion: http://postgr.es/m/CA+Tgmobp7+7kmi4gkq7Y+4AM9fTvL+O1oQ4-5gFTT+6Ng-dQ=g@mail.gmail.com
Discussion: http://postgr.es/m/CAFiTN-vTe79M8uDH1yprOU64MNFE+R3ODRuA+JWf27JbhY4hJw@mail.gmail.com
There were many calls to construct_array() and deconstruct_array() for
built-in types, for example, when dealing with system catalog columns.
These all hardcoded the type attributes necessary to pass to these
functions.
To simplify this a bit, add construct_array_builtin(),
deconstruct_array_builtin() as wrappers that centralize this hardcoded
knowledge. This simplifies many call sites and reduces the amount of
hardcoded stuff that is spread around.
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://www.postgresql.org/message-id/flat/2914356f-9e5f-8c59-2995-5997fc48bcba%40enterprisedb.com
When rebuilding the relation mapping on subscribers, we were not releasing
the attribute mapping's memory which was no longer required.
The attribute mapping used in logical tuple conversion was refactored in
PG13 (by commit e1551f96e6) but we forgot to update the related code that
frees the attribute map.
Author: Hou Zhijie
Reviewed-by: Amit Langote, Amit Kapila, Shi yu
Backpatch-through: 10, where it was introduced
Discussion: https://postgr.es/m/OSZPR01MB6310F46CD425A967E4AEF736FDA49@OSZPR01MB6310.jpnprd01.prod.outlook.com
We build the partition map entries on subscribers while applying the
changes for update/delete on partitions. The component relation in each
entry is closed after its use so we need to update it on successive use of
cache entries.
This problem was there since the original commit f1ac27bfda that
introduced this code but we didn't notice it till the recent commit
26b3455afa started to use the component relation of partition map cache
entry.
Reported-by: Tom Lane, as per buildfarm
Author: Amit Langote, Hou Zhijie
Reviewed-by: Amit Kapila, Shi Yu
Backpatch-through: 13, where it was introduced
Discussion: https://postgr.es/m/OSZPR01MB6310F46CD425A967E4AEF736FDA49@OSZPR01MB6310.jpnprd01.prod.outlook.com
In logical replication, we will check if the target table on the
subscriber is updatable by comparing the replica identity of the table on
the publisher with the table on the subscriber. When the target table is a
partitioned table, we only check its replica identity but not for the
partition tables. This leads to assertion failure while applying changes
for update/delete as we expect those to succeed only when the
corresponding partition table has a primary key or has a replica
identity defined.
Fix it by checking the replica identity of the partition table while
applying changes.
Reported-by: Shi Yu
Author: Shi Yu, Hou Zhijie
Reviewed-by: Amit Langote, Amit Kapila
Backpatch-through: 13, where it was introduced
Discussion: https://postgr.es/m/OSZPR01MB6310F46CD425A967E4AEF736FDA49@OSZPR01MB6310.jpnprd01.prod.outlook.com
We were not updating the partition map cache in the subscriber even when
the corresponding remote rel is changed. Due to this data was getting
incorrectly replicated for partition tables after the publisher has
changed the table schema.
Fix it by resetting the required entries in the partition map cache after
receiving a new relation mapping from the publisher.
Reported-by: Shi Yu
Author: Shi Yu, Hou Zhijie
Reviewed-by: Amit Langote, Amit Kapila
Backpatch-through: 13, where it was introduced
Discussion: https://postgr.es/m/OSZPR01MB6310F46CD425A967E4AEF736FDA49@OSZPR01MB6310.jpnprd01.prod.outlook.com
While building a new attrmap which maps partition attribute numbers to
remoterel's, we incorrectly update the map for dropped column attributes.
Later, it caused cache look-up failure when we tried to use the map to
fetch the information about attributes.
This also fixes the partition map cache invalidation which was using the
wrong type cast to fetch the entry. We were using stale partition map
entry after invalidation which leads to the assertion or cache look-up
failure.
Reported-by: Shi Yu
Author: Hou Zhijie, Shi Yu
Reviewed-by: Amit Langote, Amit Kapila
Backpatch-through: 13, where it was introduced
Discussion: https://postgr.es/m/OSZPR01MB6310F46CD425A967E4AEF736FDA49@OSZPR01MB6310.jpnprd01.prod.outlook.com
pg_stat_get_subscription scanned one more LogicalRepWorker array entry
than is really allocated. In the worst case this could lead to SIGSEGV,
if the LogicalRepCtx data structure is near the end of shared memory.
That seems quite unlikely though (thanks to the ordering of calls in
CreateSharedMemoryAndSemaphores) and we've heard no field reports of it.
A more likely misbehavior is one row of garbage data in the function's
result, but even that is not real likely because of the check that the
pid field matches some live backend.
Report and fix by Kuntal Ghosh. This bug is old, so back-patch
to all supported branches.
Discussion: https://postgr.es/m/CAGz5QCJykEDzW6jQK6Yz7Qh_PMtD=95de_7QoocbVR2Qy8hWZA@mail.gmail.com
Currently, we simply combine the column lists when publishing tables on
multiple publications and that can sometimes lead to unexpected behavior.
Say, if a column is published in any row-filtered publication, then the
values for that column are sent to the subscriber even for rows that don't
match the row filter, as long as the row matches the row filter for any
other publication, even if that other publication doesn't include the
column.
The main purpose of introducing a column list is to have statically
different shapes on publisher and subscriber or hide sensitive column
data. In both cases, it doesn't seem to make sense to combine column
lists.
So, we disallow the cases where the column list is different for the same
table when combining publications. It can be later extended to combine the
column lists for selective cases where required.
Reported-by: Alvaro Herrera
Author: Hou Zhijie
Reviewed-by: Amit Kapila
Discussion: https://postgr.es/m/202204251548.mudq7jbqnh7r@alvherre.pgsql
Commit 923def9a53 and 52e4f0cd47 allowed to specify column lists and row
filters for publication tables. This commit extends the
pg_publication_tables view and pg_get_publication_tables function to
display that information.
This information will be useful to users and we also need this for the
later commit that prohibits combining multiple publications with different
column lists for the same table.
Author: Hou Zhijie
Reviewed By: Amit Kapila, Alvaro Herrera, Shi Yu, Takamichi Osumi
Discussion: https://postgr.es/m/202204251548.mudq7jbqnh7r@alvherre.pgsql
pgstat_report_stat() is only supposed to be called outside of transactions. In
5891c7a8ed I added a pgstat_report_stat() call into LogicalRepApplyLoop()'s
timeout branch. While not commonly reached inside a transaction, it is
reachable (e.g. due to network bottlenecks or the sender being stalled / slow
for some reason).
To fix, add a !IsTransactionState() check.
No test added because there's no easy way to reproduce this case without
patching the code.
Reported-By: Erik Rijkers <er@xs4all.nl>
Discussion: https://postgr.es/m/b3463b8c-2328-dcac-0136-af95715493c1@xs4all.nl
The problem is that we don't send keep-alive messages for a long time
while processing large transactions during logical replication where we
don't send any data of such transactions. This can happen when the table
modified in the transaction is not published or because all the changes
got filtered. We do try to send the keep_alive if necessary at the end of
the transaction (via WalSndWriteData()) but by that time the
subscriber-side can timeout and exit.
To fix this we try to send the keepalive message if required after
processing certain threshold of changes.
Reported-by: Fabrice Chapuis
Author: Wang wei and Amit Kapila
Reviewed By: Masahiko Sawada, Euler Taveira, Hou Zhijie, Hayato Kuroda
Backpatch-through: 10
Discussion: https://postgr.es/m/CAA5-nLARN7-3SLU_QUxfy510pmrYK6JJb=bk3hcgemAM_pAv+w@mail.gmail.com
Allow extensions to specify a new custom resource manager (rmgr),
which allows specialized WAL. This is meant to be used by a Table
Access Method or Index Access Method.
Prior to this commit, only Generic WAL was available, which offers
support for recovery and physical replication but not logical
replication.
Reviewed-by: Julien Rouhaud, Bharath Rupireddy, Andres Freund
Discussion: https://postgr.es/m/ed1fb2e22d15d3563ae0eb610f7b61bb15999c0a.camel%40j-davis.com
Previously the statistics collector received statistics updates via UDP and
shared statistics data by writing them out to temporary files regularly. These
files can reach tens of megabytes and are written out up to twice a
second. This has repeatedly prevented us from adding additional useful
statistics.
Now statistics are stored in shared memory. Statistics for variable-numbered
objects are stored in a dshash hashtable (backed by dynamic shared
memory). Fixed-numbered stats are stored in plain shared memory.
The header for pgstat.c contains an overview of the architecture.
The stats collector is not needed anymore, remove it.
By utilizing the transactional statistics drop infrastructure introduced in a
prior commit statistics entries cannot "leak" anymore. Previously leaked
statistics were dropped by pgstat_vacuum_stat(), called from [auto-]vacuum. On
systems with many small relations pgstat_vacuum_stat() could be quite
expensive.
Now that replicas drop statistics entries for dropped objects, it is not
necessary anymore to reset stats when starting from a cleanly shut down
replica.
Subsequent commits will perform some further code cleanup, adapt docs and add
tests.
Bumps PGSTAT_FILE_FORMAT_ID.
Author: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Author: Andres Freund <andres@anarazel.de>
Author: Melanie Plageman <melanieplageman@gmail.com>
Reviewed-By: Andres Freund <andres@anarazel.de>
Reviewed-By: Thomas Munro <thomas.munro@gmail.com>
Reviewed-By: Justin Pryzby <pryzby@telsasoft.com>
Reviewed-By: "David G. Johnston" <david.g.johnston@gmail.com>
Reviewed-By: Tomas Vondra <tomas.vondra@2ndquadrant.com> (in a much earlier version)
Reviewed-By: Arthur Zakirov <a.zakirov@postgrespro.ru> (in a much earlier version)
Reviewed-By: Antonin Houska <ah@cybertec.at> (in a much earlier version)
Discussion: https://postgr.es/m/20220303021600.hs34ghqcw6zcokdh@alap3.anarazel.de
Discussion: https://postgr.es/m/20220308205351.2xcn6k4x5yivcxyd@alap3.anarazel.de
Discussion: https://postgr.es/m/20210319235115.y3wz7hpnnrshdyv6@alap3.anarazel.de
Previously the pgstat <-> replication slots API was done with on the basis of
names. However, the upcoming move to storing stats in shared memory makes it
more convenient to use a integer as key.
Change the replication slot functions to take the slot rather than the slot
name, and expose ReplicationSlotIndex() to compute the index of an replication
slot. Special handling will be required for restarts, as the index is not
stable across restarts. For now pgstat internally still uses names.
Rename pgstat_report_replslot_{create,drop}() to
pgstat_{create,drop}_replslot() to match the functions for other kinds of
stats.
Reviewed-By: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Discussion: https://postgr.es/m/20220404041516.cctrvpadhuriawlq@alap3.anarazel.de
Here we make a series of improvements to the generation memory
allocator, namely:
1. Allow generation contexts to have a minimum, initial and maximum block
sizes. The standard allocator allows this already but when the generation
context was added, it only allowed fixed-sized blocks. The problem with
fixed-sized blocks is that it's difficult to choose how large to make the
blocks. If the chosen size is too small then we'd end up with a large
number of blocks and a large number of malloc calls. If the block size is
made too large, then memory is wasted.
2. Add support for "keeper" blocks. This is a special block that is
allocated along with the context itself but is never freed. Instead,
when the last chunk in the keeper block is freed, we simply mark the block
as empty to allow new allocations to make use of it.
3. Add facility to "recycle" newly empty blocks instead of freeing them
and having to later malloc an entire new block again. We do this by
recording a single GenerationBlock which has become empty of any chunks.
When we run out of space in the current block, we check to see if there is
a "freeblock" and use that if it contains enough space for the allocation.
Author: David Rowley, Tomas Vondra
Reviewed-by: Andy Fan
Discussion: https://postgr.es/m/d987fd54-01f8-0f73-af6c-519f799a0ab8@enterprisedb.com
The current logical replication behavior is to send every transaction to
subscriber even if the transaction is empty. This can happen because
transaction doesn't contain changes from the selected publications or all
the changes got filtered. It is a waste of CPU cycles and network
bandwidth to build/transmit these empty transactions.
This patch addresses the above problem by postponing the BEGIN message
until the first change is sent. While processing a COMMIT message, if
there was no other change for that transaction, do not send the COMMIT
message. This allows us to skip sending BEGIN/COMMIT messages for empty
transactions.
When skipping empty transactions in synchronous replication mode, we send
a keepalive message to avoid delaying such transactions.
Author: Ajin Cherian, Hou Zhijie, Euler Taveira
Reviewed-by: Peter Smith, Takamichi Osumi, Shi Yu, Masahiko Sawada, Greg Nancarrow, Vignesh C, Amit Kapila
Discussion: https://postgr.es/m/CAMkU=1yohp9-dv48FLoSPrMqYEyyS5ZWkaZGD41RJr10xiNo_Q@mail.gmail.com
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
This commit adds support for decoding of sequences to the built-in
replication (the infrastructure was added by commit 0da92dc530).
The syntax and behavior mostly mimics handling of tables, i.e. a
publication may be defined as FOR ALL SEQUENCES (replicating all
sequences in a database), FOR ALL SEQUENCES IN SCHEMA (replicating
all sequences in a particular schema) or individual sequences.
To publish sequence modifications, the publication has to include
'sequence' action. The protocol is extended with a new message,
describing sequence increments.
A new system view pg_publication_sequences lists all the sequences
added to a publication, both directly and indirectly. Various psql
commands (\d and \dRp) are improved to also display publications
including a given sequence, or sequences included in a publication.
Author: Tomas Vondra, Cary Huang
Reviewed-by: Peter Eisentraut, Amit Kapila, Hannu Krosing, Andres
Freund, Petr Jelinek
Discussion: https://postgr.es/m/d045f3c2-6cfb-06d3-5540-e63c320df8bc@enterprisedb.com
Discussion: https://postgr.es/m/1710ed7e13b.cd7177461430746.3372264562543607781@highgo.ca
This feature allows skipping the transaction on subscriber nodes.
If incoming change violates any constraint, logical replication stops
until it's resolved. Currently, users need to either manually resolve the
conflict by updating a subscriber-side database or by using function
pg_replication_origin_advance() to skip the conflicting transaction. This
commit introduces a simpler way to skip the conflicting transactions.
The user can specify LSN by ALTER SUBSCRIPTION ... SKIP (lsn = XXX),
which allows the apply worker to skip the transaction finished at
specified LSN. The apply worker skips all data modification changes within
the transaction.
Author: Masahiko Sawada
Reviewed-by: Takamichi Osumi, Hou Zhijie, Peter Eisentraut, Amit Kapila, Shi Yu, Vignesh C, Greg Nancarrow, Haiying Tang, Euler Taveira
Discussion: https://postgr.es/m/CAD21AoDeScrsHhLyEPYqN3sydg6PxAPVBboK=30xJfUVihNZDA@mail.gmail.com
Teach xlogreader.c to decode the WAL into a circular buffer. This will
support optimizations based on looking ahead, to follow in a later
commit.
* XLogReadRecord() works as before, decoding records one by one, and
allowing them to be examined via the traditional XLogRecGetXXX()
macros and certain traditional members like xlogreader->ReadRecPtr.
* An alternative new interface XLogReadAhead()/XLogNextRecord() is
added that returns pointers to DecodedXLogRecord objects so that it's
now possible to look ahead in the WAL stream while replaying.
* In order to be able to use the new interface effectively while
streaming data, support is added for the page_read() callback to
respond to a new nonblocking mode with XLREAD_WOULDBLOCK instead of
waiting for more data to arrive.
No direct user of the new interface is included in this commit, though
XLogReadRecord() uses it internally. Existing code doesn't need to
change, except in a few places where it was accessing reader internals
directly and now needs to go through accessor macros.
Reviewed-by: Julien Rouhaud <rjuju123@gmail.com>
Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com>
Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions)
Discussion: https://postgr.es/m/CA+hUKGJ4VJN8ttxScUFM8dOKX0BrBiboo5uz1cq=AovOddfHpA@mail.gmail.com
Logical replication apply workers for a subscription can easily get stuck
in an infinite loop of attempting to apply a change, triggering an error
(such as a constraint violation), exiting with the error written to the
subscription server log, and restarting.
To partially remedy the situation, this patch adds a new subscription
option named 'disable_on_error'. To be consistent with old behavior, this
option defaults to false. When true, both the tablesync worker and apply
worker catch any errors thrown and disable the subscription in order to
break the loop. The error is still also written in the logs.
Once the subscription is disabled, users can either manually resolve the
conflict/error or skip the conflicting transaction by using
pg_replication_origin_advance() function. After resolving the conflict,
users need to enable the subscription to allow apply process to proceed.
Author: Osumi Takamichi and Mark Dilger
Reviewed-by: Greg Nancarrow, Vignesh C, Amit Kapila, Wang wei, Tang Haiying, Peter Smith, Masahiko Sawada, Shi Yu
Discussion : https://postgr.es/m/DB35438F-9356-4841-89A0-412709EBD3AB%40enterprisedb.com
This commits adds both the finish LSN (commit_lsn in case transaction got
committed, prepare_lsn in case of a prepared transaction, etc.) and
replication origin name to the existing error context message.
This will help users in specifying the origin name and transaction finish
LSN to pg_replication_origin_advance() SQL function to skip a particular
transaction.
Author: Masahiko Sawada
Reviewed-by: Takamichi Osumi, Euler Taveira, and Amit Kapila
Discussion: https://postgr.es/m/CAD21AoBarBf2oTF71ig2g_o=3Z_Dt6_sOpMQma1kFgbnA5OZ_w@mail.gmail.com
Previously, the message for logical replication worker errcontext is
incrementally built, which was not translation friendly. Instead, we use
complete sentences with if-else branches.
We also remove the commit timestamp from the context message since it's
not important information and made the message long.
Author: Masahiko Sawada
Reviewed-by: Takamichi Osumi, and Amit Kapila
Discussion: https://postgr.es/m/CAD21AoBarBf2oTF71ig2g_o=3Z_Dt6_sOpMQma1kFgbnA5OZ_w@mail.gmail.com
Set-returning functions that use the Materialize mode, creating a
tuplestore to include all the tuples returned in a set rather than doing
so in multiple calls, use roughly the same set of steps to prepare
ReturnSetInfo for this job:
- Check if ReturnSetInfo supports returning a tuplestore and if the
materialize mode is enabled.
- Create a tuplestore for all the tuples part of the returned set in the
per-query memory context, stored in ReturnSetInfo->setResult.
- Build a tuple descriptor mostly from get_call_result_type(), then
stored in ReturnSetInfo->setDesc. Note that there are some cases where
the SRF's tuple descriptor has to be the one specified by the function
caller.
This refactoring is done so as there are (well, should be) no behavior
changes in any of the in-core functions refactored, and the centralized
function that checks and sets up the function's ReturnSetInfo can be
controlled with a set of bits32 options. Two of them prove to be
necessary now:
- SRF_SINGLE_USE_EXPECTED to use expectedDesc as tuple descriptor, as
expected by the function's caller.
- SRF_SINGLE_BLESS to validate the tuple descriptor for the SRF.
The same initialization pattern is simplified in 28 places per my
count as of src/backend/, shaving up to ~900 lines of code. These
mostly come from the removal of the per-query initializations and the
sanity checks now grouped in a single location. There are more
locations that could be simplified in contrib/, that are left for a
follow-up cleanup.
fcc2817, 07daca5 and d61a361 have prepared the areas of the code related
to this change, to ease this refactoring.
Author: Melanie Plageman, Michael Paquier
Reviewed-by: Álvaro Herrera, Justin Pryzby
Discussion: https://postgr.es/m/CAAKRu_azyd1Z3W_r7Ou4sorTjRCs+PxeHw1CWJeXKofkE6TuZg@mail.gmail.com
It was decided (refer to the Discussion link below) that the stats
collector is not an appropriate place to store the error information of
subscription workers.
This patch changes the pg_stat_subscription_workers view (introduced by
commit 8d74fc96db) so that it stores only statistics counters:
apply_error_count and sync_error_count, and has one entry for
each subscription. The removed error information such as error-XID and
the error message would be stored in another way in the future which is
more reliable and persistent.
After removing these error details, there is no longer any relation
information, so the subscription statistics are now a cluster-wide
statistics.
The patch also changes the view name to pg_stat_subscription_stats since
the word "worker" is an implementation detail that we use one worker for
one tablesync and one apply.
Author: Masahiko Sawada, based on suggestions by Andres Freund
Reviewed-by: Peter Smith, Haiying Tang, Takamichi Osumi, Amit Kapila
Discussion: https://postgr.es/m/20220125063131.4cmvsxbz2tdg6g65@alap3.anarazel.de
This feature adds row filtering for publication tables. When a publication
is defined or modified, an optional WHERE clause can be specified. Rows
that don't satisfy this WHERE clause will be filtered out. This allows a
set of tables to be partially replicated. The row filter is per table. A
new row filter can be added simply by specifying a WHERE clause after the
table name. The WHERE clause must be enclosed by parentheses.
The row filter WHERE clause for a table added to a publication that
publishes UPDATE and/or DELETE operations must contain only columns that
are covered by REPLICA IDENTITY. The row filter WHERE clause for a table
added to a publication that publishes INSERT can use any column. If the
row filter evaluates to NULL, it is regarded as "false". The WHERE clause
only allows simple expressions that don't have user-defined functions,
user-defined operators, user-defined types, user-defined collations,
non-immutable built-in functions, or references to system columns. These
restrictions could be addressed in the future.
If you choose to do the initial table synchronization, only data that
satisfies the row filters is copied to the subscriber. If the subscription
has several publications in which a table has been published with
different WHERE clauses, rows that satisfy ANY of the expressions will be
copied. If a subscriber is a pre-15 version, the initial table
synchronization won't use row filters even if they are defined in the
publisher.
The row filters are applied before publishing the changes. If the
subscription has several publications in which the same table has been
published with different filters (for the same publish operation), those
expressions get OR'ed together so that rows satisfying any of the
expressions will be replicated.
This means all the other filters become redundant if (a) one of the
publications have no filter at all, (b) one of the publications was
created using FOR ALL TABLES, (c) one of the publications was created
using FOR ALL TABLES IN SCHEMA and the table belongs to that same schema.
If your publication contains a partitioned table, the publication
parameter publish_via_partition_root determines if it uses the partition's
row filter (if the parameter is false, the default) or the root
partitioned table's row filter.
Psql commands \dRp+ and \d <table-name> will display any row filters.
Author: Hou Zhijie, Euler Taveira, Peter Smith, Ajin Cherian
Reviewed-by: Greg Nancarrow, Haiying Tang, Amit Kapila, Tomas Vondra, Dilip Kumar, Vignesh C, Alvaro Herrera, Andres Freund, Wei Wang
Discussion: https://www.postgresql.org/message-id/flat/CAHE3wggb715X%2BmK_DitLXF25B%3DjE6xyNCH4YOwM860JR7HarGQ%40mail.gmail.com
This routine is a no-op since dd04e95 from 2003, with a macro kept
around for compatibility purposes. This has led to the same code
patterns being copy-pasted around for no effect, sometimes in confusing
ways like in pg_logical_slot_get_changes_guts() from logical.c where the
code was actually incorrect.
This issue has been discussed on two different threads recently, so
rather than living with this legacy, remove any uses of this routine in
the C code to simplify things. The compatibility macro is kept to avoid
breaking any out-of-core modules that depend on it.
Reported-by: Tatsuhito Kasahara, Justin Pryzby
Author: Tatsuhito Kasahara
Discussion: https://postgr.es/m/20211217200419.GQ17618@telsasoft.com
Discussion: https://postgr.es/m/CAP0=ZVJeeYfAeRfmzqAF2Lumdiv4S4FewyBnZd4DPTrsSQKJKw@mail.gmail.com
This moves the functions related to performing WAL recovery into the new
xlogrecovery.c source file, leaving xlog.c responsible for maintaining
the WAL buffers, coordinating the startup and switch from recovery to
normal operations, and other miscellaneous stuff that have always been in
xlog.c.
Reviewed-by: Andres Freund, Kyotaro Horiguchi, Robert Haas
Discussion: https://www.postgresql.org/message-id/a31f27b4-a31d-f976-6217-2b03be646ffa%40iki.fi
This extends the logical decoding to also decode sequence increments.
We differentiate between sequences created in the current (in-progress)
transaction, and sequences created earlier. This mixed behavior is
necessary because while sequences are not transactional (increments are
not subject to ROLLBACK), relfilenode changes are. So we do this:
* Changes for sequences created in the same top-level transaction are
treated as transactional, i.e. just like any other change from that
transaction, and discarded in case of a rollback.
* Changes for sequences created earlier are applied immediately, as if
performed outside any transaction. This applies also after ALTER
SEQUENCE, which may create a new relfilenode.
Moreover, if we ever get support for DDL replication, the sequence
won't exist until the transaction gets applied.
Sequences created in the current transaction are tracked in a simple
hash table, identified by a relfilenode. That means a sequence may
already exist, but if a transaction does ALTER SEQUENCE then the
increments for the new relfilenode will be treated as transactional.
For each relfilenode we track the XID of (sub)transaction that created
it, which is needed for cleanup at transaction end. We don't need to
check the XID to decide if an increment is transactional - if we find a
match in the hash table, it has to be the same transaction.
This requires two minor changes to WAL-logging. Firstly, we need to
ensure the sequence record has a valid XID - until now the the increment
might have XID 0 if it was the first change in a subxact. But the
sequence might have been created in the same top-level transaction. So
we ensure the XID is assigned when WAL-logging increments.
The other change is addition of "created" flag, marking increments for
newly created relfilenodes. This makes it easier to maintain the hash
table of sequences that need transactional handling.
Note: This is needed because of subxacts. A XID 0 might still have the
sequence created in a different subxact of the same top-level xact.
This does not include any changes to test_decoding and/or the built-in
replication - those will be committed in separate patches.
A patch adding decoding of sequences was originally submitted by Cary
Huang. This commit reworks various important aspects (e.g. the WAL
logging and transactional/non-transactional handling). However, the
original patch and reviews were very useful.
Author: Tomas Vondra, Cary Huang
Reviewed-by: Peter Eisentraut, Hannu Krosing, Andres Freund
Discussion: https://postgr.es/m/d045f3c2-6cfb-06d3-5540-e63c320df8bc@enterprisedb.com
Discussion: https://postgr.es/m/1710ed7e13b.cd7177461430746.3372264562543607781@highgo.ca
xlog.h is directly and indirectly #included in a lot of places. With
this change, xloginsert.h is no longer unnecessarily included in the
large number of them that don't need it.
Author: Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>
Discussion: https://postgr.es/m/CALj2ACVe-W+WM5P44N7eG9C2_FmaeM8Dq5aCnD3fHt0Ba=WR6w@mail.gmail.com
This fixes a set of issues that have accumulated over the past months
(or years) in various code areas. Most fixes are related to some recent
additions, as of the development of v15.
Author: Justin Pryzby
Discussion: https://postgr.es/m/20220124030001.GQ23027@telsasoft.com
Commit 5c649fe15 introduced a memory leak into pg_basebackup's
parse_compress_options. (I simplified nearby code while at it.)
Commit 9a974cbcb introduced a memory leak into pg_dump's
binary_upgrade_set_pg_class_oids.
Coverity also complained about a call of SnapBuildProcessChange that
ignored the result, unlike every other call of that function. This
is evidently intentional, so add a (void) cast to indicate that.
(It's also old, dating to b89e15105; I suppose the reason it showed
up now is 7a5f6b474's recent rearrangement of nearby code.)
Prevent logical replication workers from performing insert, update,
delete, truncate, or copy commands on tables unless the subscription
owner has permission to do so.
Prevent subscription owners from circumventing row-level security by
forbidding replication into tables with row-level security policies
which the subscription owner is subject to, without regard to whether
the policy would ordinarily allow the INSERT, UPDATE, DELETE or
TRUNCATE which is being replicated. This seems sufficient for now, as
superusers, roles with bypassrls, and target table owners should still
be able to replicate despite RLS policies. We can revisit the
question of applying row-level security policies on a per-row basis if
this restriction proves too severe in practice.
Author: Mark Dilger
Reviewed-by: Jeff Davis, Andrew Dunstan, Ronan Dunklau
Discussion: https://postgr.es/m/9DFC88D3-1300-4DE8-ACBC-4CEF84399A53%40enterprisedb.com
This commit adds a new system view pg_stat_subscription_workers, that
shows information about any errors which occur during the application of
logical replication changes as well as during performing initial table
synchronization. The subscription statistics entries are removed when the
corresponding subscription is removed.
It also adds an SQL function pg_stat_reset_subscription_worker() to reset
single subscription errors.
The contents of this view can be used by an upcoming patch that skips the
particular transaction that conflicts with the existing data on the
subscriber.
This view can be extended in the future to track other xact related
statistics like the number of xacts committed/aborted for subscription
workers.
Author: Masahiko Sawada
Reviewed-by: Greg Nancarrow, Hou Zhijie, Tang Haiying, Vignesh C, Dilip Kumar, Takamichi Osumi, Amit Kapila
Discussion: https://postgr.es/m/CAD21AoDeScrsHhLyEPYqN3sydg6PxAPVBboK=30xJfUVihNZDA@mail.gmail.com
A couple of spots can benefit from an added errdetail(), which matches
what we were already doing in other places; and those that cannot
withstand errdetail() can get a more descriptive primary message.
Author: Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>
Reviewed-by: Daniel Gustafsson <daniel@yesql.se>
Reviewed-by: Julien Rouhaud <rjuju123@gmail.com>
Discussion: https://postgr.es/m/CALj2ACV+cX1eM03GfcA=ZMLXh5fSn1X1auJLz3yuS1duPSb9QA@mail.gmail.com
protocol.sgml documented the layout for Type messages, but completely
dropped the ball otherwise, failing to explain what they are, when
they are sent, or what they're good for. While at it, do a little
copy-editing on the description of Relation messages.
In passing, adjust the comment for apply_handle_type() to make it
clearer that we choose not to do anything when receiving a Type
message, not that we think it has no use whatsoever.
Per question from Stefen Hillman.
Discussion: https://postgr.es/m/CAPgW8pMknK5pup6=T4a_UG=Cz80Rgp=KONqJmTdHfaZb0RvnFg@mail.gmail.com
A couple of code paths related to logical decoding (WAL sender, slot
advancing, etc.) use XLogReadRecord(), feeding on error messages
generated by walreader.c on a failure. All those messages have no
context, making it harder to spot from where an error could come even if
these should not happen. All the other callers of XLogReadRecord() do
that already.
Reviewed-by: Kyotaro Horiguchi
Discussion: https://postgr.es/m/YYnTH6OyOwQcAdkw@paquier.xyz
All such code deals with this global variable in one of three ways.
Sometimes the same functions use it in more than one of these ways
at the same time.
First, sometimes it's an implicit argument to one or more functions
being called in xlog.c or elsewhere, and must be set to the
appropriate value before calling those functions lest they
misbehave. In those cases, it is now passed as an explicit argument
instead.
Second, sometimes it's used to obtain the current timeline after
the end of recovery, i.e. the timeline to which WAL is being
written and flushed. Such code now calls GetWALInsertionTimeLine()
or relies on the new out parameter added to GetFlushRecPtr().
Third, sometimes it's used during recovery to store the current
replay timeline. That can change, so such code must generally
update the value before each use. It can still do that, but must
now use a local variable instead.
The net effect of these changes is to reduce by a fair amount the
amount of code that is directly accessing this global variable.
That's good, because history has shown that we don't always think
clearly about which timeline ID it's supposed to contain at any
given point in time, or indeed, whether it has been or needs to
be initialized at any given point in the code.
Patch by me, reviewed and tested by Michael Paquier, Amul Sul, and
Álvaro Herrera.
Discussion: https://postgr.es/m/CA+TgmobfAAqhfWa1kaFBBFvX+5CjM=7TE=n4r4Q1o2bjbGYBpA@mail.gmail.com
In slotfuncs.c, pg_replication_slot_advance() needs to determine
the LSN up to which the slot should be advanced, but that doesn't
require us to update ThisTimeLineID, because none of the code called
from here depends on it. If the replication slot is logical,
pg_logical_replication_slot_advance will call read_local_xlog_page,
which does use ThisTimeLineID, but also takes care of making sure
it's up to date. If the replication slot is physical, the timeline
isn't used for anything at all.
In logicalfuncs.c, pg_logical_slot_get_changes_guts() has the same
issue: the only code we're going to run that cares about timelines
is in or downstream of read_local_xlog_page, which already makes
sure that the correct value gets set. Hence, don't do it here.
Patch by me, reviewed and tested by Michael Paquier, Amul Sul, and
Álvaro Herrera.
Discussion: https://postgr.es/m/CA+TgmobfAAqhfWa1kaFBBFvX+5CjM=7TE=n4r4Q1o2bjbGYBpA@mail.gmail.com
Failing to do so results in inability of logical decoding to process the
WAL stream. Handle it by doing nothing.
Backpatch all the way back.
Reported-by: Petr Jelínek <petr.jelinek@enterprisedb.com>
During a replication slot creation, an ERROR generated in the same
transaction as the one creating a to-be-exported snapshot would have
left the backend in an inconsistent state, as the associated static
export snapshot state was not being reset on transaction abort, but only
on the follow-up command received by the WAL sender that created this
snapshot on replication slot creation. This would trigger inconsistency
failures if this session tried to export again a snapshot, like during
the creation of a replication slot.
Note that a snapshot export cannot happen in a transaction block, so
there is no need to worry resetting this state for subtransaction
aborts. Also, this inconsistent state would very unlikely show up to
users. For example, one case where this could happen is an
out-of-memory error when building the initial snapshot to-be-exported.
Dilip found this problem while poking at a different patch, that caused
an error in this code path for reasons unrelated to HEAD.
Author: Dilip Kumar
Reviewed-by: Michael Paquier, Zhihong Yu
Discussion: https://postgr.es/m/CAFiTN-s0zA1Kj0ozGHwkYkHwa5U0zUE94RSc_g81WrpcETB5=w@mail.gmail.com
Backpatch-through: 9.6
Two functions, both named check_permissions(), used the same checks to
verify if a user had required privileges to work on replication slots.
This commit removes the duplication, and moves the function doing the
checks to slot.c to be centralized.
Author: Bharath Rupireddy
Reviewed-by: Nathan Bossart, Euler Taveira
Discussion: https://postgr.es/m/CALj2ACUPpVw1u7sQocFVWrSs0n10pt_G_4NPZKSxXK6cW1dErw@mail.gmail.com
While processing toast changes in logical decoding, we rejigger the
tuple change to point to in-memory toast tuples instead to on-disk toast
tuples. And, to make sure the memory accounting is correct, we were
subtracting the old change size and then after re-computing the new tuple,
re-adding its size at the end. Now, if there is any error before we add
the new size, we will release the changes and that will update the
accounting info (subtracting the size from the counters). And we were
underflowing there which leads to an assertion failure in assert enabled
builds and wrong memory accounting in reorder buffer otherwise.
Author: Bertrand Drouvot
Reviewed-by: Amit Kapila
Backpatch-through: 13, where memory accounting was introduced
Discussion: https://postgr.es/m/92b0ee65-b8bd-e42d-c082-4f3f4bf12d34@amazon.com
Similar to LogicalIncreaseRestartDecodingForSlot() add a debug message to
LogicalIncreaseXminForSlot() reporting a new catalog_xmin candidate.
This just adds additional diagnostic information during logical decoding that
can aid debugging.
Author: Ashutosh Bapat
Reviewed-by: Masahiko Sawada, Amit Kapila
Discussion: https://postgr.es/m/CAExHW5usQWbiUz0hHOCu5twS1O9DvpcPojf6sor=8q--VUuMbA@mail.gmail.com
Use one fileset for the entire worker lifetime instead of using
separate filesets for each streaming transaction. Now, the
changes/subxacts files for every streaming transaction will be
created under the same fileset and the files will be deleted
after the transaction is completed.
This patch extends the BufFileOpenFileSet and BufFileDeleteFileSet
APIs to allow users to specify whether to give an error on missing
files.
Author: Dilip Kumar, based on suggestion by Thomas Munro
Reviewed-by: Hou Zhijie, Masahiko Sawada, Amit Kapila
Discussion: https://postgr.es/m/E1mCC6U-0004Ik-Fs@gemulon.postgresql.org
Move fileset related implementation out of sharedfileset.c to allow its
usage by backends that don't want to share filesets among different
processes. After this split, fileset infrastructure is used by both
sharedfileset.c and worker.c for the named temporary files that survive
across transactions.
Author: Dilip Kumar, based on suggestion by Andres Freund
Reviewed-by: Hou Zhijie, Masahiko Sawada, Amit Kapila
Discussion: https://postgr.es/m/E1mCC6U-0004Ik-Fs@gemulon.postgresql.org
Previously, on the subscriber, we set the error context callback for the
tuple data conversion failures. This commit replaces the existing error
context callback with a comprehensive one so that it shows not only the
details of data conversion failures but also the details of logical change
being applied by the apply worker or table sync worker. The additional
information displayed will be the command, transaction id, and timestamp.
The error context is added to an error only when applying a change but not
while doing other work like receiving data etc.
This will help users in diagnosing the problems that occur during logical
replication. It also can be used for future work that allows skipping a
particular transaction on the subscriber.
Author: Masahiko Sawada
Reviewed-by: Hou Zhijie, Greg Nancarrow, Haiying Tang, Amit Kapila
Tested-by: Haiying Tang
Discussion: https://postgr.es/m/CAD21AoDeScrsHhLyEPYqN3sydg6PxAPVBboK=30xJfUVihNZDA@mail.gmail.com
In the code, most places used the term "Stream Stop" for the logical
stream message. This commit improves consistency by renaming LogicalRepMsgType
"LOGICAL_REP_MSG_STREAM_END" to "LOGICAL_REP_MSG_STREAM_STOP".
Author: Masahiko Sawada
Reviewed-by: Hou Zhijie, Amit Kapila
Discussion: https://postgr.es/m/CAD21AoDeScrsHhLyEPYqN3sydg6PxAPVBboK=30xJfUVihNZDA@mail.gmail.com
Commit a8fd13cab0 added support for prepared transactions to built-in
logical replication via a new option "two_phase" for a subscription. The
"two_phase" option was not allowed with the existing streaming option.
This commit permits the combination of "streaming" and "two_phase"
subscription options. It extends the pgoutput plugin and the subscriber
side code to add the prepare API for streaming transactions which will
apply the changes accumulated in the spool-file at prepare time.
Author: Peter Smith and Ajin Cherian
Reviewed-by: Vignesh C, Amit Kapila, Greg Nancarrow
Tested-By: Haiying Tang
Discussion: https://postgr.es/m/02DA5F5E-CECE-4D9C-8B4B-418077E2C010@postgrespro.ru
Discussion: https://postgr.es/m/CAMGcDxeqEpWj3fTXwqhSwBdXd2RS9jzwWscO-XbeCfso6ts3+Q@mail.gmail.com
This is a non-functional change only to refactor code to extract some
replication logic into static functions.
This is done as preparation for the 2PC streaming patch which also shares
this common logic.
Author: Peter Smith
Reviewed-By: Amit Kapila
Discussion: https://postgr.es/m/CAHut+PuiSA8AiLcE2N5StzSKs46SQEP_vDOUD5fX2XCVtfZ7mQ@mail.gmail.com
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
The existing code tried to do syscache lookups in an already-failed
transaction, which is problematic to say the least. After some
consideration of alternatives, the best fix seems to be to just drop
type names from the error message altogether. The table and column
names seem like sufficient localization. If the user is unsure what
types are involved, she can check the local and remote table
definitions.
Having done that, we can also discard the LogicalRepTypMap hash
table, which had no other use. Arguably, LOGICAL_REP_MSG_TYPE
replication messages are now obsolete as well; but we should
probably keep them in case some other use emerges. (The complexity
of removing something from the replication protocol would likely
outweigh any savings anyhow.)
Masahiko Sawada and Bharath Rupireddy, per complaint from Andres
Freund. Back-patch to v10 where this code originated.
Discussion: https://postgr.es/m/20210106020229.ne5xnuu6wlondjpe@alap3.anarazel.de
Until now, we didn't allow to stream the changes in logical replication
till we receive speculative confirm or the next DML change record after
speculative inserts. The reason was that we never use to process
speculative aborts but after commit 4daa140a2f it is possible to process
them so we can allow streaming once we receive speculative abort after
speculative insertion.
We decided to backpatch to 14 where the feature for streaming in progress
transactions have been introduced as this is a minor change and makes that
functionality better.
Author: Amit Kapila
Reviewed-By: Dilip Kumar
Backpatch-through: 14
Discussion: https://postgr.es/m/CAA4eK1KdqmTCtrBR6oFfGELrLLbDLDedL6zACcsUOQuTJBj1vw@mail.gmail.com
I started out with the goal of reporting ERRCODE_CONNECTION_FAILURE
when walrcv_connect() fails, but as I looked around I realized that
whoever wrote this code was of the opinion that errcodes are purely
optional. That's not my understanding of our project policy. Hence,
make sure that an errcode is provided in each ereport that (a) is
ERROR or higher level and (b) isn't arguably an internal logic error.
Also fix some very dubious existing errcode assignments.
While this is not per policy, it's also largely cosmetic, since few
of these cases could get reported to applications. So I don't
feel a need to back-patch.
Discussion: https://postgr.es/m/2189704.1623512522@sss.pgh.pa.us
During decoding for speculative inserts, we were relying for cleaning
toast hash on confirmation records or next change records. But that
could lead to multiple problems (a) memory leak if there is neither a
confirmation record nor any other record after toast insertion for a
speculative insert in the transaction, (b) error and assertion failures
if the next operation is not an insert/update on the same table.
The fix is to start queuing spec abort change and clean up toast hash
and change record during its processing. Currently, we are queuing the
spec aborts for both toast and main table even though we perform cleanup
while processing the main table's spec abort record. Later, if we have a
way to distinguish between the spec abort record of toast and the main
table, we can avoid queuing the change for spec aborts of toast tables.
Reported-by: Ashutosh Bapat
Author: Dilip Kumar
Reviewed-by: Amit Kapila
Backpatch-through: 9.6, where it was introduced
Discussion: https://postgr.es/m/CAExHW5sPKF-Oovx_qZe4p5oM6Dvof7_P+XgsNAViug15Fm99jA@mail.gmail.com
Using an Assert to check the validity of incoming messages is an
extremely poor decision. In a debug build, it should not be that easy
for a broken or malicious remote client to crash the logrep worker.
The consequences could be even worse in non-debug builds, which will
fail to make such checks at all, leading to who-knows-what misbehavior.
Hence, promote every Assert that could possibly be triggered by wrong
or out-of-order replication messages to a full test-and-ereport.
To avoid bloating the set of messages the translation team has to cope
with, establish a policy that replication protocol violation error
reports don't need to be translated. Hence, all the new messages here
use errmsg_internal(). A couple of old messages are changed likewise
for consistency.
Along the way, fix some non-idiomatic or outright wrong uses of
hash_search().
Most of these mistakes are new with the "streaming replication"
patch (commit 464824323), but a couple go back a long way.
Back-patch as appropriate.
Discussion: https://postgr.es/m/1719083.1623351052@sss.pgh.pa.us
apply_handle_tuple_routing(), having detected and reported that
the tuple it needed to update didn't exist, tried to update that
tuple anyway, leading to a null-pointer dereference.
logicalrep_partition_open() failed to ensure that the
LogicalRepPartMapEntry it built for a partition was fully
independent of that for the partition root, leading to
trouble if the root entry was later freed or rebuilt.
Meanwhile, on the publisher's side, pgoutput_change() sometimes
attempted to apply execute_attr_map_tuple() to a NULL tuple.
The first of these was reported by Sergey Bernikov in bug #17055;
I found the other two while developing some test cases for this
sadly under-tested code.
Diagnosis and patch for the first issue by Amit Langote; patches
for the others by me; new test cases by me. Back-patch to v13
where this logic came in.
Discussion: https://postgr.es/m/17055-9ba800ec8522668b@postgresql.org
It turns out that worker.c's code path for TRUNCATE was also
careless about establishing a snapshot while executing user-defined
code, allowing the checks added by commit 84f5c2908 to fail when
a trigger is fired in that context.
We could just wrap Push/PopActiveSnapshot around the truncate call,
but it seems better to establish a policy of holding a snapshot
throughout execution of a replication step. To help with that and
possible future requirements, replace the previous ensure_transaction
calls with pairs of begin/end_replication_step calls.
Per report from Mark Dilger. Back-patch to v11, like the previous
changes.
Discussion: https://postgr.es/m/B4A3AF82-79ED-4F4C-A4E5-CD2622098972@enterprisedb.com
While decoding the multi-insert WAL we can't clean the toast untill we get
the last insert of that WAL record. Now if we stream the changes before we
get the last change, the memory for toast chunks won't be released and we
expect the txn to have streamed all changes after streaming. This
restriction is mainly to ensure the correctness of streamed transactions
and it doesn't seem worth uplifting such a restriction just to allow this
case because anyway we will stream the transaction once such an insert is
complete.
Previously we were using two different flags (one for toast tuples and
another for speculative inserts) to indicate partial changes. Now instead
we replaced both of them with a single flag to indicate partial changes.
Reported-by: Pavan Deolasee
Author: Dilip Kumar
Reviewed-by: Pavan Deolasee, Amit Kapila
Discussion: https://postgr.es/m/CABOikdN-_858zojYN-2tNcHiVTw-nhxPwoQS4quExeweQfG1Ug@mail.gmail.com
If we redirected a replicated tuple operation into a partition child
table, and then tried to fire AFTER triggers for that event, the
relation cache entry for the child table was already closed. This has
no visible ill effects as long as the entry is still there and still
valid, but an unluckily-timed cache flush could result in a crash or
other misbehavior.
To fix, postpone the ExecCleanupTupleRouting call (which is what
closes the child table) until after we've fired triggers. This
requires a bit of refactoring so that the cleanup function can
have access to the necessary state.
In HEAD, I took the opportunity to simplify some of worker.c's
function APIs based on use of the new ApplyExecutionData struct.
However, it doesn't seem safe/practical to back-patch that aspect,
at least not without a lot of analysis of possible interactions
with a04daa97a.
In passing, add an Assert to afterTriggerInvokeEvents to catch
such cases. This seems worthwhile because we've grown a number
of fairly unstructured ways of calling AfterTriggerEndQuery.
Back-patch to v13, where worker.c grew the ability to deal with
partitioned target tables.
Discussion: https://postgr.es/m/3382681.1621381328@sss.pgh.pa.us
COMMIT/ROLLBACK necessarily destroys all snapshots within the session.
The original implementation of intra-procedure transactions just
cavalierly did that, ignoring the fact that this left us executing in
a rather different environment than normal. In particular, it turns
out that handling of toasted datums depends rather critically on there
being an outer ActiveSnapshot: otherwise, when SPI or the core
executor pop whatever snapshot they used and return, it's unsafe to
dereference any toasted datums that may appear in the query result.
It's possible to demonstrate "no known snapshots" and "missing chunk
number N for toast value" errors as a result of this oversight.
Historically this outer snapshot has been held by the Portal code,
and that seems like a good plan to preserve. So add infrastructure
to pquery.c to allow re-establishing the Portal-owned snapshot if it's
not there anymore, and add enough bookkeeping support that we can tell
whether it is or not.
We can't, however, just re-establish the Portal snapshot as part of
COMMIT/ROLLBACK. As in normal transaction start, acquiring the first
snapshot should wait until after SET and LOCK commands. Hence, teach
spi.c about doing this at the right time. (Note that this patch
doesn't fix the problem for any PLs that try to run intra-procedure
transactions without using SPI to execute SQL commands.)
This makes SPI's no_snapshots parameter rather a misnomer, so in HEAD,
rename that to allow_nonatomic.
replication/logical/worker.c also needs some fixes, because it wasn't
careful to hold a snapshot open around AFTER trigger execution.
That code doesn't use a Portal, which I suspect someday we're gonna
have to fix. But for now, just rearrange the order of operations.
This includes back-patching the recent addition of finish_estate()
to centralize the cleanup logic there.
This also back-patches commit 2ecfeda3e into v13, to improve the
test coverage for worker.c (it was that test that exposed that
worker.c's snapshot management is wrong).
Per bug #15990 from Andreas Wicht. Back-patch to v11 where
intra-procedure COMMIT was added.
Discussion: https://postgr.es/m/15990-eee2ac466b11293d@postgresql.org
While applying the truncate change, the logical apply worker acquires
RowExclusiveLock on the relation being truncated. This allowed truncate on
the relation at a time by two apply workers which lead to a deadlock. The
reason was that one of the workers after updating the pg_class tuple tries
to acquire SHARE lock on the relation and started to wait for the second
worker which has acquired RowExclusiveLock on the relation. And when the
second worker tries to update the pg_class tuple, it starts to wait for
the first worker which leads to a deadlock. Fix it by acquiring
AccessExclusiveLock on the relation before applying the truncate change as
we do for normal truncate operation.
Author: Peter Smith, test case by Haiying Tang
Reviewed-by: Dilip Kumar, Amit Kapila
Backpatch-through: 11
Discussion: https://postgr.es/m/CAHut+PsNm43p0jM+idTvWwiGZPcP0hGrHMPK9TOAkc+a4UpUqw@mail.gmail.com
The worker.c global wrconn is only meant to be used by logical apply/
tablesync workers, but there are other variables with the same name. To
reduce future confusion rename the global from "wrconn" to
"LogRepWorkerWalRcvConn".
While this is just cosmetic, it seems better to backpatch it all the way
back to 10 where this code appeared, to avoid future backpatching
issues.
Author: Peter Smith <smithpb2250@gmail.com>
Discussion: https://postgr.es/m/CAHut+Pu7Jv9L2BOEx_Z0UtJxfDevQSAUW2mJqWU+CtmDrEZVAg@mail.gmail.com
Also "make reformat-dat-files".
The only change worthy of note is that pgindent messed up the formatting
of launcher.c's struct LogicalRepWorkerId, which led me to notice that
that struct wasn't used at all anymore, so I just took it out.
This set of commits has some bugs with known fixes, but at this late
stage in the release cycle it seems best to revert and resubmit next
time, along with some new automated test coverage for this whole area.
Commits reverted:
dc88460c: Doc: Review for "Optionally prefetch referenced data in recovery."
1d257577: Optionally prefetch referenced data in recovery.
f003d9f8: Add circular WAL decoding buffer.
323cbe7c: Remove read_page callback from XLogReader.
Remove the new GUC group WAL_RECOVERY recently added by a55a9847, as the
corresponding section of config.sgml is now reverted.
Discussion: https://postgr.es/m/CAOuzzgrn7iKnFRsB4MHp3UisEQAGgZMbk_ViTN4HV4-Ksq8zCg%40mail.gmail.com
Currently, replication slot statistics are updated at prepare, commit, and
rollback. Now, if the transaction is interrupted the stats might not get
updated. Fixed this by updating replication statistics after every
stream/spill.
In passing update the docs to change the description of some of the slot
stats.
Author: Vignesh C, Sawada Masahiko
Reviewed-by: Amit Kapila
Discussion: https://postgr.es/m/20210319185247.ldebgpdaxsowiflw@alap3.anarazel.de
During decoding of an in-progress or prepared transaction, we detect
concurrent abort with an error code ERRCODE_TRANSACTION_ROLLBACK. That is
not sufficient because a callback can decide to throw that error code
at other times as well.
Reported-by: Tom Lane
Author: Amit Kapila
Reviewed-by: Dilip Kumar
Discussion: https://postgr.es/m/CAA4eK1KCjPRS4aZHB48QMM4J8XOC1+TD8jo-4Yu84E+MjwqVhA@mail.gmail.com
Previously, we were using the size of all the changes present in
ReorderBuffer to compute total_bytes after decoding a transaction and that
can lead to counting some of the transactions' changes more than once. Fix
it by using the size of the changes decoded for a transaction to compute
'total_bytes'.
Author: Sawada Masahiko
Reviewed-by: Vignesh C, Amit Kapila
Discussion: https://postgr.es/m/20210319185247.ldebgpdaxsowiflw@alap3.anarazel.de
There were two problems:
a. We were always selecting the next available txn instead of selecting it
when it is larger than the previous transaction.
b. We were selecting the transactions which haven't made any changes to
the database (base snapshot is not set). Later it was hitting an Assert
because we don't decode such transactions and the changes in txn remain as
it is. It is better not to choose such transactions for streaming in the
first place.
Reported-by: Haiying Tang
Author: Dilip Kumar
Reviewed-by: Amit Kapila
Discussion: https://postgr.es/m/OS0PR01MB61133B94E63177040F7ECDA1FB429@OS0PR01MB6113.jpnprd01.prod.outlook.com
Commit 8ff1c94649 allowed TRUNCATE command to truncate foreign tables.
Previously the information about "ONLY" options specified in TRUNCATE
command were passed to the foreign data wrapper. Then postgres_fdw
constructed the TRUNCATE command to issue the remote server and
included "ONLY" options in it based on the passed information.
On the other hand, "ONLY" options specified in SELECT, UPDATE or DELETE
have no effect when accessing or modifying the remote table, i.e.,
are not passed to the foreign data wrapper. So it's inconsistent to
make only TRUNCATE command pass the "ONLY" options to the foreign data
wrapper. Therefore this commit changes the TRUNCATE command so that
it doesn't pass the "ONLY" options to the foreign data wrapper,
for the consistency with other statements. Also this commit changes
postgres_fdw so that it always doesn't include "ONLY" options in
the TRUNCATE command that it constructs.
Author: Fujii Masao
Reviewed-by: Bharath Rupireddy, Kyotaro Horiguchi, Justin Pryzby, Zhihong Yu
Discussion: https://postgr.es/m/551ed8c1-f531-818b-664a-2cecdab99cd8@oss.nttdata.com
Previously, we used to use the array of size max_replication_slots to
store stats for replication slots. But that had two problems in the cases
where a message for dropping a slot gets lost: 1) the stats for the new
slot are not recorded if the array is full and 2) writing beyond the end
of the array if the user reduces the max_replication_slots.
This commit uses HTAB for replication slot statistics, resolving both
problems. Now, pgstat_vacuum_stat() search for all the dead replication
slots in stats hashtable and tell the collector to remove them. To avoid
showing the stats for the already-dropped slots, pg_stat_replication_slots
view searches slot stats by the slot name taken from pg_replication_slots.
Also, we send a message for creating a slot at slot creation, initializing
the stats. This reduces the possibility that the stats are accumulated
into the old slot stats when a message for dropping a slot gets lost.
Reported-by: Andres Freund
Author: Sawada Masahiko, test case by Vignesh C
Reviewed-by: Amit Kapila, Vignesh C, Dilip Kumar
Discussion: https://postgr.es/m/20210319185247.ldebgpdaxsowiflw@alap3.anarazel.de
The Truncate operation acquires an exclusive lock on the target relation
and indexes. It then waits for logical replication of the operation to
finish at commit. Now because we are acquiring the shared lock on the
target index to get index attributes in pgoutput while sending the
changes for the Truncate operation, it leads to a deadlock.
Actually, we don't need to acquire a lock on the target index as we build
the cache entry using a historic snapshot and all the later changes are
absorbed while decoding WAL. So, we wrote a special purpose function for
logical replication to get a bitmap of replica identity attribute numbers
where we get that information without locking the target index.
We decided not to backpatch this as there doesn't seem to be any field
complaint about this issue since it was introduced in commit 5dfd1e5a in
v11.
Reported-by: Haiying Tang
Author: Takamichi Osumi, test case by Li Japin
Reviewed-by: Amit Kapila, Ajin Cherian
Discussion: https://postgr.es/m/OS0PR01MB6113C2499C7DC70EE55ADB82FB759@OS0PR01MB6113.jpnprd01.prod.outlook.com
We send the prepare for the concurrently aborted xacts so that later when
rollback prepared is decoded and sent, the downstream should be able to
rollback such a xact. For 'streaming' case (when we send changes for
in-progress transactions), we were sending prepare twice when concurrent
abort was detected.
Author: Peter Smith
Reviewed-by: Amit Kapila
Discussion: https://postgr.es/m/f82133c6-6055-b400-7922-97dae9f2b50b@enterprisedb.com
Creating a trigger on a relation to which an apply operation is
triggered would cause a relation leak once the change gets committed,
as the executor would miss that the relation needs to be closed
beforehand. This issue got introduced with the refactoring done in
1375422c, where it becomes necessary to track relations within
es_opened_result_relations to make sure that they are closed.
We have discussed using ExecInitResultRelation() coupled with
ExecCloseResultRelations() for the relations in need of tracking by the
apply operations in the subscribers, which would simplify greatly the
opening and closing of indexes, but this requires a larger rework and
reorganization of the worker code, particularly for the tuple routing
part. And that's not really welcome post feature freeze. So, for now,
settle down to the same solution as TRUNCATE which is to fill in
es_opened_result_relations with the relation opened, to make sure that
ExecGetTriggerResultRel() finds them and that they get closed.
The code is lightly refactored so as a relation is not registered three
times for each DML code path, making the whole a bit easier to follow.
Reported-by: Tang Haiying, Shi Yu, Hou Zhijie
Author: Amit Langote, Masahiko Sawada, Hou Zhijie
Reviewed-by: Amit Kapila, Michael Paquier
Discussion: https://postgr.es/m/OS0PR01MB611383FA0FE92EB9DE21946AFB769@OS0PR01MB6113.jpnprd01.prod.outlook.com
This adds the statistics about total transactions count and total
transaction data logically sent to the decoding output plugin from
ReorderBuffer. Users can query the pg_stat_replication_slots view to check
these stats.
Suggested-by: Andres Freund
Author: Vignesh C and Amit Kapila
Reviewed-by: Sawada Masahiko, Amit Kapila
Discussion: https://postgr.es/m/20210319185247.ldebgpdaxsowiflw@alap3.anarazel.de
This will make it consistent with the other usage of slotname in the code.
In the passing, change pgstat_report_replslot signature to use a structure
rather than multiple parameters.
Reported-by: Andres Freund
Author: Vignesh C
Reviewed-by: Sawada Masahiko, Amit Kapila
Discussion: https://postgr.es/m/20210319185247.ldebgpdaxsowiflw@alap3.anarazel.de
This commit introduces new foreign data wrapper API for TRUNCATE.
It extends TRUNCATE command so that it accepts foreign tables as
the targets to truncate and invokes that API. Also it extends postgres_fdw
so that it can issue TRUNCATE command to foreign servers, by adding
new routine for that TRUNCATE API.
The information about options specified in TRUNCATE command, e.g.,
ONLY, CACADE, etc is passed to FDW via API. The list of foreign tables to
truncate is also passed to FDW. FDW truncates the foreign data sources
that the passed foreign tables specify, based on those information.
For example, postgres_fdw constructs TRUNCATE command using them
and issues it to the foreign server.
For performance, TRUNCATE command invokes the FDW routine for
TRUNCATE once per foreign server that foreign tables to truncate belong to.
Author: Kazutaka Onishi, Kohei KaiGai, slightly modified by Fujii Masao
Reviewed-by: Bharath Rupireddy, Michael Paquier, Zhihong Yu, Alvaro Herrera, Stephen Frost, Ashutosh Bapat, Amit Langote, Daniel Gustafsson, Ibrar Ahmed, Fujii Masao
Discussion: https://postgr.es/m/CAOP8fzb_gkReLput7OvOK+8NHgw-RKqNv59vem7=524krQTcWA@mail.gmail.com
Discussion: https://postgr.es/m/CAJuF6cMWDDqU-vn_knZgma+2GMaout68YUgn1uyDnexRhqqM5Q@mail.gmail.com
Teach xlogreader.c to decode its output into a circular buffer, to
support optimizations based on looking ahead.
* XLogReadRecord() works as before, consuming records one by one, and
allowing them to be examined via the traditional XLogRecGetXXX()
macros.
* An alternative new interface XLogNextRecord() is added that returns
pointers to DecodedXLogRecord structs that can be examined directly.
* XLogReadAhead() provides a second cursor that lets you see
further ahead, as long as data is available and there is enough space
in the decoding buffer. This returns DecodedXLogRecord pointers to the
caller, but also adds them to a queue of records that will later be
consumed by XLogNextRecord()/XLogReadRecord().
The buffer's size is controlled with wal_decode_buffer_size. The buffer
could potentially be placed into shared memory, for future projects.
Large records that don't fit in the circular buffer are called
"oversized" and allocated separately with palloc().
Discussion: https://postgr.es/m/CA+hUKGJ4VJN8ttxScUFM8dOKX0BrBiboo5uz1cq=AovOddfHpA@mail.gmail.com
Previously, the XLogReader module would fetch new input data using a
callback function. Redesign the interface so that it tells the caller
to insert more data with a special return value instead. This API suits
later patches for prefetching, encryption and maybe other future
projects that would otherwise require continually extending the callback
interface.
As incidental cleanup work, move global variables readOff, readLen and
readSegNo inside XlogReaderState.
Author: Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>
Author: Heikki Linnakangas <hlinnaka@iki.fi> (parts of earlier version)
Reviewed-by: Antonin Houska <ah@cybertec.at>
Reviewed-by: Alvaro Herrera <alvherre@2ndquadrant.com>
Reviewed-by: Takashi Menjo <takashi.menjo@gmail.com>
Reviewed-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Thomas Munro <thomas.munro@gmail.com>
Discussion: https://postgr.es/m/20190418.210257.43726183.horiguchi.kyotaro%40lab.ntt.co.jp
Arrange to do some things on-demand, rather than immediately during
executor startup, because there's a fair chance of never having to do
them at all:
* Don't open result relations' indexes until needed.
* Don't initialize partition tuple routing, nor the child-to-root
tuple conversion map, until needed.
This wins in UPDATEs on partitioned tables when only some of the
partitions will actually receive updates; with larger partition
counts the savings is quite noticeable. Also, we can remove some
sketchy heuristics in ExecInitModifyTable about whether to set up
tuple routing.
Also, remove execPartition.c's private hash table tracking which
partitions were already opened by the ModifyTable node. Instead
use the hash added to ModifyTable itself by commit 86dc90056.
To allow lazy computation of the conversion maps, we now set
ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it
only in some, not terribly well-defined, cases. This has user-visible
side effects in that now more error messages refer to the root
relation instead of some partition (and provide error data in the
root's column order, too). It looks to me like this is a strict
improvement in consistency, so I don't have a problem with the
output changes visible in this commit.
Extracted from a larger patch, which seemed to me to be too messy
to push in one commit.
Amit Langote, reviewed at different times by Heikki Linnakangas and
myself
Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
Retry the call to heap_prune_page() in rare cases where there is
disagreement between the heap_prune_page() call and the call to
HeapTupleSatisfiesVacuum() that immediately follows. Disagreement is
possible when a concurrently-aborted transaction makes a tuple DEAD
during the tiny window between each step. This was the only case where
a tuple considered DEAD by VACUUM still had storage following pruning.
VACUUM's definition of dead tuples is now uniformly simple and
unambiguous: dead tuples from each page are always LP_DEAD line pointers
that were encountered just after we performed pruning (and just before
we considered freezing remaining items with tuple storage).
Eliminating the tupgone=true special case enables INDEX_CLEANUP=off
style skipping of index vacuuming that takes place based on flexible,
dynamic criteria. The INDEX_CLEANUP=off case had to know about skipping
indexes up-front before now, due to a subtle interaction with the
special case (see commit dd695979) -- this was a special case unto
itself. Now there are no special cases. And so now it won't matter
when or how we decide to skip index vacuuming: it won't affect how
pruning behaves, and it won't be affected by any of the implementation
details of pruning or freezing.
Also remove XLOG_HEAP2_CLEANUP_INFO records. These are no longer
necessary because we now rely entirely on heap pruning taking care of
recovery conflicts. There is no longer any need to generate recovery
conflicts for DEAD tuples that pruning just missed. This also means
that heap vacuuming now uses exactly the same strategy for recovery
conflicts as index vacuuming always has: REDO routines never need to
process a latestRemovedXid from the WAL record, since earlier REDO of
the WAL record from pruning is sufficient in all cases. The generic
XLOG_HEAP2_CLEAN record type is now split into two new record types to
reflect this new division (these are called XLOG_HEAP2_PRUNE and
XLOG_HEAP2_VACUUM).
Also stop acquiring a super-exclusive lock for heap pages when they're
vacuumed during VACUUM's second heap pass. A regular exclusive lock is
enough. This is correct because heap page vacuuming is now strictly a
matter of setting the LP_DEAD line pointers to LP_UNUSED. No other
backend can have a pointer to a tuple located in a pinned buffer that
can be invalidated by a concurrent heap page vacuum operation.
Heap vacuuming can now be thought of as conceptually similar to index
vacuuming and conceptually dissimilar to heap pruning. Heap pruning now
has sole responsibility for anything involving the logical contents of
the database (e.g., managing transaction status information, recovery
conflicts, considering what to do with HOT chains). Index vacuuming and
heap vacuuming are now only concerned with recycling garbage items from
physical data structures that back the logical database.
Bump XLOG_PAGE_MAGIC due to pruning and heap page vacuum WAL record
changes.
Credit for the idea of retrying pruning a page to avoid the tupgone case
goes to Andres Freund.
Author: Peter Geoghegan <pg@bowt.ie>
Reviewed-By: Andres Freund <andres@anarazel.de>
Reviewed-By: Masahiko Sawada <sawada.mshk@gmail.com>
Discussion: https://postgr.es/m/CAH2-WznneCXTzuFmcwx_EyRQgfsfJAAsu+CsqRFmFXCAar=nJw@mail.gmail.com
The output plugin accepts a new parameter (messages) that controls if
logical decoding messages are written into the replication stream. It is
useful for those clients that use pgoutput as an output plugin and needs
to process messages that were written by pg_logical_emit_message().
Although logical streaming replication protocol supports logical
decoding messages now, logical replication does not use this feature yet.
Author: David Pirotte, Euler Taveira
Reviewed-by: Euler Taveira, Andres Freund, Ashutosh Bapat, Amit Kapila
Discussion: https://postgr.es/m/CADK3HHJ-+9SO7KuRLH=9Wa1rAo60Yreq1GFNkH_kd0=CdaWM+A@mail.gmail.com
It is possible that while decoding a prepared transaction, it gets aborted
concurrently via a ROLLBACK PREPARED command. In that case, we were
skipping all the changes and directly sending Rollback Prepared when we
find the same in WAL. However, the downstream has no idea of the GID of
such a transaction. So, ensure to send prepare even when a concurrent
abort is detected.
Author: Ajin Cherian
Reviewed-by: Markus Wanner, Amit Kapila
Discussion: https://postgr.es/m/f82133c6-6055-b400-7922-97dae9f2b50b@enterprisedb.com
Put the remote end's error message into the primary error string,
instead of relegating it to errdetail(). Although this could end up
being awkward if the remote sends us a really long error message,
it seems more in keeping with our message style guidelines, and more
helpful in situations where the errdetail could get dropped.
Peter Smith
Discussion: https://postgr.es/m/CAHut+Ps-Qv2yQceCwobQDP0aJOkfDzRFrOaR6+2Op2K=WHGeWg@mail.gmail.com
Along with gid, this provides a different way to identify the transaction.
The users that use xid in some way to prepare the transactions can use it
to filter prepare transactions. The later commands COMMIT PREPARED or
ROLLBACK PREPARED carries both identifiers, providing an output plugin the
choice of what to use.
Author: Markus Wanner
Reviewed-by: Vignesh C, Amit Kapila
Discussion: https://postgr.es/m/ee280000-7355-c4dc-e47b-2436e7be959c@enterprisedb.com
There is now a per-column COMPRESSION option which can be set to pglz
(the default, and the only option in up until now) or lz4. Or, if you
like, you can set the new default_toast_compression GUC to lz4, and
then that will be the default for new table columns for which no value
is specified. We don't have lz4 support in the PostgreSQL code, so
to use lz4 compression, PostgreSQL must be built --with-lz4.
In general, TOAST compression means compression of individual column
values, not the whole tuple, and those values can either be compressed
inline within the tuple or compressed and then stored externally in
the TOAST table, so those properties also apply to this feature.
Prior to this commit, a TOAST pointer has two unused bits as part of
the va_extsize field, and a compessed datum has two unused bits as
part of the va_rawsize field. These bits are unused because the length
of a varlena is limited to 1GB; we now use them to indicate the
compression type that was used. This means we only have bit space for
2 more built-in compresison types, but we could work around that
problem, if necessary, by introducing a new vartag_external value for
any further types we end up wanting to add. Hopefully, it won't be
too important to offer a wide selection of algorithms here, since
each one we add not only takes more coding but also adds a build
dependency for every packager. Nevertheless, it seems worth doing
at least this much, because LZ4 gets better compression than PGLZ
with less CPU usage.
It's possible for LZ4-compressed datums to leak into composite type
values stored on disk, just as it is for PGLZ. It's also possible for
LZ4-compressed attributes to be copied into a different table via SQL
commands such as CREATE TABLE AS or INSERT .. SELECT. It would be
expensive to force such values to be decompressed, so PostgreSQL has
never done so. For the same reasons, we also don't force recompression
of already-compressed values even if the target table prefers a
different compression method than was used for the source data. These
architectural decisions are perhaps arguable but revisiting them is
well beyond the scope of what seemed possible to do as part of this
project. However, it's relatively cheap to recompress as part of
VACUUM FULL or CLUSTER, so this commit adjusts those commands to do
so, if the configured compression method of the table happens not to
match what was used for some column value stored therein.
Dilip Kumar. The original patches on which this work was based were
written by Ildus Kurbangaliev, and those were patches were based on
even earlier work by Nikita Glukhov, but the design has since changed
very substantially, since allow a potentially large number of
compression methods that could be added and dropped on a running
system proved too problematic given some of the architectural issues
mentioned above; the choice of which specific compression method to
add first is now different; and a lot of the code has been heavily
refactored. More recently, Justin Przyby helped quite a bit with
testing and reviewing and this version also includes some code
contributions from him. Other design input and review from Tomas
Vondra, Álvaro Herrera, Andres Freund, Oleg Bartunov, Alexander
Korotkov, and me.
Discussion: http://postgr.es/m/20170907194236.4cefce96%40wp.localdomain
Discussion: http://postgr.es/m/CAFiTN-uUpX3ck%3DK0mLEk-G_kUQY%3DSNOTeqdaNRR9FMdQrHKebw%40mail.gmail.com
Commit ce0fdbfe97 made tablesync slots permanent and allow Drop
Subscription to drop such slots. However, it is possible that before
tablesync worker could get the acknowledgment of slot creation, drop
subscription stops it and that can lead to a dangling slot on the
publisher. Prevent cancel/die interrupts while creating a slot in the
tablesync worker.
Reported-by: Thomas Munro as per buildfarm
Author: Amit Kapila
Reviewed-by: Vignesh C, Takamichi Osumi
Discussion: https://postgr.es/m/CA+hUKGJG9dWpw1cOQ2nzWU8PHjm=PTraB+KgE5648K9nTfwvxg@mail.gmail.com
Commit 0aa8a01d04 extends the output plugin API to allow decoding of
prepared xacts and allowed the user to enable/disable the two-phase option
via pg_logical_slot_get_changes(). This can lead to a problem such that
the first time when it gets changes via pg_logical_slot_get_changes()
without two_phase option enabled it will not get the prepared even though
prepare is after consistent snapshot. Now next time during getting changes,
if the two_phase option is enabled it can skip prepare because by that
time start decoding point has been moved. So the user will only get commit
prepared.
Allow to enable/disable this option at the create slot time and default
will be false. It will break the existing slots which is fine in a major
release.
Author: Ajin Cherian
Reviewed-by: Amit Kapila and Vignesh C
Discussion: https://postgr.es/m/d0f60d60-133d-bf8d-bd70-47784d8fabf3@enterprisedb.com
In commit a271a1b50e, we allowed decoding at prepare time and the prepare
was decoded again if there is a restart after decoding it. It was done
that way because we can't distinguish between the cases where we have not
decoded the prepare because it was prior to consistent snapshot or we have
decoded it earlier but restarted. To distinguish between these two cases,
we have introduced an initial_consistent_point at the slot level which is
an LSN at which we found a consistent point at the time of slot creation.
This is also the point where we have exported a snapshot for the initial
copy. So, prepare transaction prior to this point are sent along with
commit prepared.
This commit bumps SNAPBUILD_VERSION because of change in SnapBuild. It
will break existing slots which is fine in a major release.
Author: Ajin Cherian, based on idea by Andres Freund
Reviewed-by: Amit Kapila and Vignesh C
Discussion: https://postgr.es/m/d0f60d60-133d-bf8d-bd70-47784d8fabf3@enterprisedb.com
Commit a271a1b50e introduced decoding at prepare time in ReorderBuffer.
This can lead to deadlock for out-of-core logical replication solutions
that uses this feature to build distributed 2PC in case such transactions
lock [user] catalog tables exclusively. They need to inform users to not
have locks on catalog tables (via explicit LOCK command) in such
transactions.
Reported-by: Andres Freund
Discussion: https://postgr.es/m/20210222222847.tpnb6eg3yiykzpky@alap3.anarazel.de
We don't have anything to decode in a transaction if ReorderBufferTXN
doesn't exist by the time we decode the commit prepared. So don't create a
new ReorderBufferTXN here. This is an oversight in commit a271a1b5.
Reported-by: Markus Wanner
Discussion: https://postgr.es/m/dbec82e2-dbd7-95a2-c6b6-e488cbbdf853@bluegap.ch
In 955a684e04 we fixed a bug in initial snapshot creation. In the
course of which several members of struct SnapBuild were obsoleted. As
SnapBuild is serialized to disk we couldn't change the memory layout.
Unfortunately I subsequently forgot about removing the backward compat
gunk, but luckily Heikki just reminded me.
This commit bumps SNAPBUILD_VERSION, therefore breaking existing
slots (which is fine in a major release).
Author: Andres Freund
Reminded-By: Heikki Linnakangas <hlinnaka@iki.fi>
Discussion: https://postgr.es/m/c94be044-818f-15e3-1ad3-7a7ae2dfed0a@iki.fi
For the initial table data synchronization in logical replication, we use
a single transaction to copy the entire table and then synchronize the
position in the stream with the main apply worker.
There are multiple downsides of this approach: (a) We have to perform the
entire copy operation again if there is any error (network breakdown,
error in the database operation, etc.) while we synchronize the WAL
position between tablesync worker and apply worker; this will be onerous
especially for large copies, (b) Using a single transaction in the
synchronization-phase (where we can receive WAL from multiple
transactions) will have the risk of exceeding the CID limit, (c) The slot
will hold the WAL till the entire sync is complete because we never commit
till the end.
This patch solves all the above downsides by allowing multiple
transactions during the tablesync phase. The initial copy is done in a
single transaction and after that, we commit each transaction as we
receive. To allow recovery after any error or crash, we use a permanent
slot and origin to track the progress. The slot and origin will be removed
once we finish the synchronization of the table. We also remove slot and
origin of tablesync workers if the user performs DROP SUBSCRIPTION .. or
ALTER SUBSCRIPTION .. REFERESH and some of the table syncs are still not
finished.
The commands ALTER SUBSCRIPTION ... REFRESH PUBLICATION and
ALTER SUBSCRIPTION ... SET PUBLICATION ... with refresh option as true
cannot be executed inside a transaction block because they can now drop
the slots for which we have no provision to rollback.
This will also open up the path for logical replication of 2PC
transactions on the subscriber side. Previously, we can't do that because
of the requirement of maintaining a single transaction in tablesync
workers.
Bump catalog version due to change of state in the catalog
(pg_subscription_rel).
Author: Peter Smith, Amit Kapila, and Takamichi Osumi
Reviewed-by: Ajin Cherian, Petr Jelinek, Hou Zhijie and Amit Kapila
Discussion: https://postgr.es/m/CAA4eK1KHJxaZS-fod-0fey=0tq3=Gkn4ho=8N4-5HWiCfu0H1A@mail.gmail.com
Currently, we get the origin id from the name and then drop the origin by
taking ExclusiveLock on ReplicationOriginRelationId. So, two concurrent
sessions can get the id from the name at the same time and then when they
try to drop the origin, one of the sessions will get the either
"tuple concurrently deleted" or "cache lookup failed for replication
origin ..".
To prevent this race condition we do the entire operation under lock. This
obviates the need for replorigin_drop() API and we have removed it so if
any extension authors are using it they need to instead use
replorigin_drop_by_name. See it's usage in pg_replication_origin_drop().
Author: Peter Smith
Reviewed-by: Amit Kapila, Euler Taveira, Petr Jelinek, and Alvaro
Herrera
Discussion: https://www.postgresql.org/message-id/CAHut%2BPuW8DWV5fskkMWWMqzt-x7RPcNQOtJQBp6SdwyRghCk7A%40mail.gmail.com
Up until now, we've held this lock when performing a checkpoint or
restartpoint, but commit 076a055acf back
in 2004 and commit 7e48b77b1c from 2009,
taken together, have removed all need for this. In the present code,
there's only ever one process entitled to attempt a checkpoint: either
the checkpointer, during normal operation, or the postmaster, during
single-user operation. So, we don't need the lock.
One possible concern in making this change is that it means that
a substantial amount of code where HOLD_INTERRUPTS() was previously
in effect due to the preceding LWLockAcquire() will now be
running without that. This could mean that ProcessInterrupts()
gets called in places from which it didn't before. However, this
seems unlikely to do very much, because the checkpointer doesn't
have any signal mapped to die(), so it's not clear how,
for example, ProcDiePending = true could happen in the first
place. Similarly with ClientConnectionLost and recovery conflicts.
Also, if there are any such problems, we might want to fix them
rather than reverting this, since running lots of code with
interrupt handling suspended is generally bad.
Patch by me, per an inquiry by Amul Sul. Review by Tom Lane
and Michael Paquier.
Discussion: http://postgr.es/m/CAAJ_b97XnBBfYeSREDJorFsyoD1sHgqnNuCi=02mNQBUMnA=FA@mail.gmail.com
Add an executor aminsert() hint mechanism that informs index AMs that
the incoming index tuple (the tuple that accompanies the hint) is not
being inserted by execution of an SQL statement that logically modifies
any of the index's key columns.
The hint is received by indexes when an UPDATE takes place that does not
apply an optimization like heapam's HOT (though only for indexes where
all key columns are logically unchanged). Any index tuple that receives
the hint on insert is expected to be a duplicate of at least one
existing older version that is needed for the same logical row. Related
versions will typically be stored on the same index page, at least
within index AMs that apply the hint.
Recognizing the difference between MVCC version churn duplicates and
true logical row duplicates at the index AM level can help with cleanup
of garbage index tuples. Cleanup can intelligently target tuples that
are likely to be garbage, without wasting too many cycles on less
promising tuples/pages (index pages with little or no version churn).
This is infrastructure for an upcoming commit that will teach nbtree to
perform bottom-up index deletion. No index AM actually applies the hint
just yet.
Author: Peter Geoghegan <pg@bowt.ie>
Reviewed-By: Victor Yegorov <vyegorov@gmail.com>
Discussion: https://postgr.es/m/CAH2-Wz=CEKFa74EScx_hFVshCOn6AA5T-ajFASTdzipdkLTNQQ@mail.gmail.com