Our documentation states that our maximum field size is 1 GB, and that
our maximum row size of 1.6 TB. However, while this might be attainable
in theory with enough contortions, it is not workable in practice; for
starters, pg_dump fails to dump tables containing rows larger than 1 GB,
even if individual columns are well below the limit; and even if one
does manage to manufacture a dump file containing a row that large, the
server refuses to load it anyway.
This commit enables dumping and reloading of such tuples, provided two
conditions are met:
1. no single column is larger than 1 GB (in output size -- for bytea
this includes the formatting overhead)
2. the whole row is not larger than 2 GB
There are three related changes to enable this:
a. StringInfo's API now has two additional functions that allow creating
a string that grows beyond the typical 1GB limit (and "long" string).
ABI compatibility is maintained. We still limit these strings to 2 GB,
though, for reasons explained below.
b. COPY now uses long StringInfos, so that pg_dump doesn't choke
trying to emit rows longer than 1GB.
c. heap_form_tuple now uses the MCXT_ALLOW_HUGE flag in its allocation
for the input tuple, which means that large tuples are accepted on
input. Note that at this point we do not apply any further limit to the
input tuple size.
The main reason to limit to 2 GB is that the FE/BE protocol uses 32 bit
length words to describe each row; and because the documentation is
ambiguous on its signedness and libpq does consider it signed, we cannot
use the highest-order bit. Additionally, the StringInfo API uses "int"
(which is 4 bytes wide in most platforms) in many places, so we'd need
to change that API too in order to improve, which has lots of fallout.
Backpatch to 9.5, which is the oldest that has
MemoryContextAllocExtended, a necessary piece of infrastructure. We
could apply to 9.4 with very minimal additional effort, but any further
than that would require backpatching "huge" allocations too.
This is the largest set of changes we could find that can be
back-patched without breaking compatibility with existing systems.
Fixing a bigger set of problems (for example, dumping tuples bigger than
2GB, or dumping fields bigger than 1GB) would require changing the FE/BE
protocol and/or changing the StringInfo API in an ABI-incompatible way,
neither of which would be back-patchable.
Authors: Daniel Vérité, Álvaro Herrera
Reviewed by: Tomas Vondra
Discussion: https://postgr.es/m/20160229183023.GA286012@alvherre.pgsql
The whole walreceiver API is now wrapped into a struct, like most of our
other loadable module APIs. The libpq connection is no longer a global
variable in libpqwalreceiver. Instead, it is encapsulated into a struct
that is passed around the functions. This allows multiple walreceivers
to run at the same time.
Add some rudimentary support for logical replication connections to
libpqwalreceiver.
These changes are mostly cosmetic and are going to be useful for the
future logical replication patches.
From: Petr Jelinek <petr@2ndquadrant.com>
Replace use of poll()/select() by WaitLatchOrSocket(), which is more
portable and flexible.
Also change walreceiver to use its procLatch instead of a custom latch.
From: Petr Jelinek <petr@2ndquadrant.com>
This works like function_with_argtypes, but aggregates allow slightly
different arguments.
Reviewed-by: Alvaro Herrera <alvherre@2ndquadrant.com>
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
It was apparently added for use by GRANT/REVOKE, but move it closer to
where other function signature related things are kept.
Reviewed-by: Alvaro Herrera <alvherre@2ndquadrant.com>
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Instead of sometimes referring to a function signature like func_name
func_args, use the existing function_with_argtypes symbol, which
combines the two.
Reviewed-by: Alvaro Herrera <alvherre@2ndquadrant.com>
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
So far the hashtable stored representative tuples in the form of its
input slot, with all columns in the hashtable that are not
needed (i.e. not grouped upon or functionally dependent) set to NULL.
Thats good for saving memory, but it turns out that having tuples full
of NULL isn't free. slot_deform_tuple is faster if there's no NULL
bitmap even if no NULLs are encountered, and skipping over leading NULLs
isn't free.
So compute a separate tuple descriptor that only contains the needed
columns. As columns have already been moved in/out the slot for the
hashtable that does not imply additional per-row overhead.
Author: Andres Freund
Reviewed-By: Heikki Linnakangas
Discussion: https://postgr.es/m/20161103110721.h5i5t5saxfk5eeik@alap3.anarazel.de
Previously we did a ExecProject() for each individual aggregate
argument. That turned out to be a performance bottleneck in queries with
multiple aggregates.
Doing all the argument computations in one ExecProject() is quite a bit
cheaper because ExecProject's fastpath can do the work at once in a
relatively tight loop, and because it can get all the required columns
with a single slot_getsomeattr and save some other redundant setup
costs.
Author: Andres Freund
Reviewed-By: Heikki Linnakangas
Discussion: https://postgr.es/m/20161103110721.h5i5t5saxfk5eeik@alap3.anarazel.de
Previously, the right to split a bucket was represented by a
heavyweight lock on the page number of the primary bucket page.
Unfortunately, this meant that every scan needed to take a heavyweight
lock on that bucket also, which was bad for concurrency. Instead, use
a cleanup lock on the primary bucket page to indicate the right to
begin a split, so that scans only need to retain a pin on that page,
which is they would have to acquire anyway, and which is also much
cheaper.
In addition to reducing the locking cost, this also avoids locking out
scans and inserts for the entire lifetime of the split: while the new
bucket is being populated with copies of the appropriate tuples from
the old bucket, scans and inserts can happen in parallel. There are
minor concurrency improvements for vacuum operations as well, though
the situation there is still far from ideal.
This patch also removes the unworldly assumption that a split will
never be interrupted. With the new code, a split is done in a series
of small steps and the system can pick up where it left off if it is
interrupted prior to completion. While this patch does not itself add
write-ahead logging for hash indexes, it is clearly a necessary first
step, since one of the things that could interrupt a split is the
removal of electrical power from the machine performing it.
Amit Kapila. I wrote the original design on which this patch is
based, and did a good bit of work on the comments and README through
multiple rounds of review, but all of the code is Amit's. Also
reviewed by Jesper Pedersen, Jeff Janes, and others.
Discussion: http://postgr.es/m/CAA4eK1LfzcZYxLoXS874Ad0+S-ZM60U9bwcyiUZx9mHZ-KCWhw@mail.gmail.com
Generate EUC_CN mappings from gb-18030-2000.xml, because GB2312.TXT is no
longer available.
Get UHC from windows-949-2000.xml, it's more up-to-date.
Plus tons more small changes. With these changes, the perl scripts
faithfully produce the *.map files we have in the repository, from the
external source files.
In the passing, fix the Makefile to also download CP932.TXT and CP950.TXT.
Based on patches by Kyotaro Horiguchi, reviewed by Daniel Gustafsson.
Discussion: https://postgr.es/m/08e7892a-d55c-eefe-76e6-7910bc8dd1f3@iki.fi
The common style is to pad to 4 digits.
Running the current perl scripts to generate these map files would override
this change, but the next commit will rewrite the perl scripts to produce
this style. I'm doing this as a separate commit, to make it more clear what
non-cosmetic changes the next commit makes to the map files.
Discussion: https://postgr.es/m/08e7892a-d55c-eefe-76e6-7910bc8dd1f3@iki.fi
PostgreSQL treats characters with < 0x80 leading byte as plain ASCII, and
they are not even passed to the conversion routines. There is no point in
having them in the conversion tables.
Everything in the tables were direct ASCII-ASCII mappings, except for two:
* SHIFT_JIS_2004 code point 0x5C (backslash in ASCII) was mapped to Unicode
YEN SIGN character.
* Unicode 0x5C (backslash again) was mapped to "REVERSE SOLIDUS" in
SHIFT_JIS_2004
These mappings never had any effect, so there's no functional change from
removing them.
Discussion: https://postgr.es/m/08e7892a-d55c-eefe-76e6-7910bc8dd1f3@iki.fi
consider_parallel_nestloop passed the wrong jointype down to its
subroutines for JOIN_UNIQUE_INNER cases (it should pass JOIN_INNER), and it
thought that it could pass paths other than innerrel->cheapest_total_path
to create_unique_path, which create_unique_path is not on board with.
These bugs would lead to assertion failures or other errors, suggesting
that this code path hasn't been tested much.
hash_inner_and_outer's code for parallel join effectively treated both
JOIN_UNIQUE_OUTER and JOIN_UNIQUE_INNER the same as JOIN_INNER (for
different reasons :-(), leading to incorrect plans that treated a semijoin
as if it were a plain join.
Michael Day submitted a test case demonstrating that hash_inner_and_outer
failed for JOIN_UNIQUE_OUTER, and I found the other cases through code
review.
Report: https://postgr.es/m/D0E8A029-D1AC-42E8-979A-5DE4A77E4413@rcmail.com
If we don't have any MCV statistics for the inner relation, and we don't
trust its numdistinct estimate either, eqjoinsel_semi falls back to a very
conservative estimate (that 50% of the outer rows have matches). This is
particularly problematic if the inner relation is completely empty, since
then even an explicit ANALYZE won't produce any pg_statistic entries,
so there's no way to budge the planner off the bad estimate.
We'd produce a better estimate in such cases if we used the nd2/nd1
selectivity heuristic, so an easy fix is to treat the nd2 estimate as
non-default if we derive it from clamping to the inner rel's rowcount
estimate. This won't fix every related case (mainly because the rowcount
estimate might be larger than DEFAULT_NUM_DISTINCT), but it seems like a
sane extension of the existing logic, so let's apply the change in HEAD
and see if anyone complains. Per bug #14438 from Nikolay Nikitin.
Report: https://postgr.es/m/20161128182113.6527.58926@wrigleys.postgresql.org
Discussion: https://postgr.es/m/31089.1480384713@sss.pgh.pa.us
Oversight in my commit 0bb51aa96. Noted while poking at a recent
bug report --- HEAD's estimates for a query using CURRENT_DATE
were unexpectedly much worse than 9.6's.
Commit a734fd5d1 exposed some race conditions that existed previously
in the autovac code, but were basically harmless because autovac would
not try to delete orphaned relations immediately. Specifically, the test
for orphaned-ness was made on a pg_class tuple that might be dead by now,
allowing autovac to try to remove a table that the owning backend had just
finished deleting. This resulted in a hard crash due to inadequate caution
about accessing the table's catalog entries without any lock. We must take
a relation lock and then recheck whether the table is still present and
still looks deletable before we do anything.
Also, it seemed to me that deleting multiple tables per transaction, and
trying to continue after errors, represented unjustifiable complexity.
We do not expect this code path to be taken often in the field, nor even
during testing, which means that prioritizing performance over correctness
is a bad tradeoff. Rip all that out in favor of just starting a new
transaction after each successful temp table deletion. If we're unlucky
enough to get an error, which shouldn't happen anyway now that we're being
more cautious, let the autovacuum worker fail as it normally would.
In passing, improve the order of operations in the initial scan loop.
Now that we don't care about whether a temp table is a wraparound hazard,
there's no need to perform extract_autovac_opts, get_pgstat_tabentry_relid,
or relation_needs_vacanalyze for temp tables.
Also, if GetTempNamespaceBackendId returns InvalidBackendId (indicating
it doesn't recognize the schema as temp), treat that as meaning it's NOT
an orphaned temp table, not that it IS one, which is what happened before
because BackendIdGetProc necessarily failed. The case really shouldn't
come up for a table that has RELPERSISTENCE_TEMP, but the consequences
if it did seem undesirable. (This might represent a back-patchable bug
fix; not sure if it's worth the trouble.)
Discussion: https://postgr.es/m/21299.1480272347@sss.pgh.pa.us
Commit 08dd23cec introduced an exception to the rule that extension member
objects can only be dropped as part of dropping the whole extension,
intending to allow such drops while running the extension's own creation or
update scripts. However, the exception was only applied at the outermost
recursion level, because it was modeled on a pre-existing check to ignore
dependencies on objects listed in pendingObjects. Bug #14434 from Philippe
Beaudoin shows that this is inadequate: in some cases we can reach an
extension member object by recursion from another one. (The bug concerns
the serial-sequence case; I'm not sure if there are other cases, but there
might well be.)
To fix, revert 08dd23cec's changes to findDependentObjects() and instead
apply the creating_extension exception regardless of stack level.
Having seen this example, I'm a bit suspicious that the pendingObjects
logic is also wrong and such cases should likewise be allowed at any
recursion level. However, changing that would interact in subtle ways
with the recursion logic (at least it would need to be moved to after the
recursing-from check). Given that the code's been like that a long time,
I'll refrain from touching it without a clear example showing it's wrong.
Back-patch to all active branches. In HEAD and 9.6, where suitable
test infrastructure exists, add a regression test case based on the
bug report.
Report: <20161125151448.6529.33039@wrigleys.postgresql.org>
Discussion: <13224.1480177514@sss.pgh.pa.us>
Instead of confusingly stating platform-dependent defaults for these
parameters in the comments in postgresql.conf.sample (with the main
entry being a lie on Linux), teach initdb to install the correct
platform-dependent value in postgresql.conf, similarly to the way
we handle other platform-dependent defaults. This won't do anything
for existing 9.6 installations, but since it's effectively only a
documentation improvement, that seems OK.
Since this requires initdb to have access to the default values,
move the #define's for those to pg_config_manual.h; the original
placement in bufmgr.h is unworkable because that file can't be
included by frontend programs.
Adjust the default value for wal_writer_flush_after so that it is 1MB
regardless of XLOG_BLCKSZ, conforming to what is stated in both the
SGML docs and postgresql.conf. (We could alternatively make it scale
with XLOG_BLCKSZ, but I'm not sure I see the point.)
Copy-edit related SGML documentation.
Fabien Coelho and Tom Lane, per a gripe from Tomas Vondra.
Discussion: <30ebc6e3-8358-09cf-44a8-578252938424@2ndquadrant.com>
Andreas Seltenreich found another case where we were being too optimistic
about allowing a plan to be considered parallelizable despite it containing
initPlans. It seems like the real issue here is that if we know we are
going to tack initPlans onto the topmost Plan node for a subquery, we
had better mark that subquery's result Paths as not-parallel-safe. That
fixes this problem and allows reversion of a kluge (added in commit
7b67a0a49 and extended in f24cf960d) to not trust the parallel_safe flag
at top level.
Discussion: <874m2w4k5d.fsf@ex.ansel.ydns.eu>
When dropping a foreign key constraint with ALTER TABLE DROP CONSTRAINT,
we refuse the drop if there are any pending trigger events on the named
table; this ensures that we won't remove the pg_trigger row that will be
consulted by those events. But we should make the same check for the
referenced relation, else we might remove a due-to-be-referenced pg_trigger
row for that relation too, resulting in "could not find trigger NNN" or
"relation NNN has no triggers" errors at commit. Per bug #14431 from
Benjie Gillam. Back-patch to all supported branches.
Report: <20161124114911.6530.31200@wrigleys.postgresql.org>
Previously, requesting commit timestamp for transactions
FrozenTransactionId and BootstrapTransactionId resulted in an error.
But since those values can validly appear in committed tuples' Xmin,
this behavior is unhelpful and error prone: each caller would have to
special-case those values before requesting timestamp data for an Xid.
We already have a perfectly good interface for returning "the Xid you
requested is too old for us to have commit TS data for it", so let's use
that instead.
Backpatch to 9.5, where commit timestamps appeared.
Author: Craig Ringer
Discussion: https://www.postgresql.org/message-id/CAMsr+YFM5Q=+ry3mKvWEqRTxrB0iU3qUSRnS28nz6FJYtBwhJg@mail.gmail.com
When rebuilding an existing index, ALTER TABLE correctly kept the
physical file in the same tablespace, but it messed up the pg_class
entry if the index had been in the database's default tablespace
and "default_tablespace" was set to some non-default tablespace.
This led to an inaccessible index.
Fix by fixing pg_get_indexdef_string() to always include a tablespace
clause, whether or not the index is in the default tablespace. The
previous behavior was installed in commit 537e92e41, and I think it just
wasn't thought through very clearly; certainly the possible effect of
default_tablespace wasn't considered. There's some risk in changing the
behavior of this function, but there are no other call sites in the core
code. Even if it's being used by some third party extension, it's fairly
hard to envision a usage that is okay with a tablespace clause being
appended some of the time but can't handle it being appended all the time.
Back-patch to all supported versions.
Code fix by me, investigation and test cases by Michael Paquier.
Discussion: <1479294998857-5930602.post@n3.nabble.com>
A new thing also called a "barrier" is proposed, but whether we decide
to take that patch or not, this file seems to have outlived its
usefulness.
Thomas Munro
Previously, the right-hand side of a multiple-column assignment, if it
wasn't a sub-SELECT, had to be a simple parenthesized expression list,
because gram.y was responsible for "bursting" the construct into
independent column assignments. This had the minor defect that you
couldn't write ROW (though you should be able to, since the standard says
this is a row constructor), and the rather larger defect that unlike other
uses of row constructors, we would not expand a "foo.*" item into multiple
columns.
Fix that by changing the RHS to be just "a_expr" in the grammar, leaving
it to transformMultiAssignRef to separate the elements of a RowExpr;
which it will do only after performing standard transformation of the
RowExpr, so that "foo.*" behaves as expected.
The key reason we didn't do that before was the hard-wired handling of
DEFAULT tokens (SetToDefault nodes). This patch deals with that issue by
allowing DEFAULT in any a_expr and having parse analysis throw an error
if SetToDefault is found in an unexpected place. That's an improvement
anyway since the error can be more specific than just "syntax error".
The SQL standard suggests that the RHS could be any a_expr yielding a
suitable row value. This patch doesn't really move the goal posts in that
respect --- you're still limited to RowExpr or a sub-SELECT --- but it does
fix the grammar restriction, so it provides some tangible progress towards
a full implementation. And the limitation is now documented by an explicit
error message rather than an unhelpful "syntax error".
Discussion: <8542.1479742008@sss.pgh.pa.us>
Condition variables provide a flexible way to sleep until a
cooperating process causes an arbitrary condition to become true. In
simple cases, this can be accomplished with a WaitLatch/ResetLatch
loop; the cooperating process can call SetLatch after performing work
that might cause the condition to be satisfied, and the waiting
process can recheck the condition each time. However, if the process
performing the work doesn't have an easy way to identify which
processes might be waiting, this doesn't work, because it can't
identify which latches to set. Condition variables solve that problem
by internally maintaining a list of waiters; a process that may have
caused some waiter's condition to be satisfied must "signal" or
"broadcast" on the condition variable.
Robert Haas and Thomas Munro
The code was intentionally not very careful about leaking strdup'd
strings in case of an error. That was forgivable probably, but it
also failed to notice strdup() failures, which could lead to subsequent
null-pointer-dereference crashes, since many callers unsurprisingly
didn't check for null pointers in the struct lconv fields. An even
worse problem is that it could throw error while we were setlocale'd
to a non-C locale, causing unwanted behavior in subsequent libc calls.
Rewrite to ensure that we cannot throw elog(ERROR) until after we've
restored the previous locale settings, or at least attempted to.
(I'm sorely tempted to make restore failure be a FATAL error, but
will refrain for the moment.) Having done that, it's not much more
work to ensure that we clean up strdup'd storage on the way out, too.
This code is substantially the same in all supported branches, so
back-patch all the way.
Michael Paquier and Tom Lane
Discussion: <CAB7nPqRMbGqa_mesopcn4MPyTs34eqtVEK7ELYxvvV=oqS00YA@mail.gmail.com>
Fix thinko in commit da1c91631: even if the original query was free of
parallel hazards, we might introduce such a hazard by adding PARAM_EXEC
Param nodes. Adjust is_parallel_safe() so that it will scan the given
expression whenever any such nodes have been created. Per report from
Andreas Seltenreich.
Discussion: <878tse6yvf.fsf@credativ.de>
Previously, we only dropped an orphan temp table when it became old
enough to threaten wraparound; instead, doing it immediately. The
only value of waiting is that someone might be able to examine the
contents of the orphan temp table for forensic purposes, but it's
pretty difficult to actually do that and few users will wish to do so.
On the flip side, not performing the drop immediately generates log
spam and bloats pg_class.
In addition, per a report from Grigory Smolkin, if a temporary schema
contains a very large number of temporary tables, a backend attempting
to clear the temporary schema might fail due to lock table exhaustion.
It's helpful for autovacuum to clean up after such cases, and we don't
want it to wait for wraparound to threaten before doing so. To
prevent autovacuum from failing in the same manner as a backend trying
to drop an entire temp schema, remove orphan temp tables in batches of
50, committing after each batch, so that we don't accumulate an
unbounded number of locks. If a drop fails, retry other orphan tables
that need to be dropped up to 10 times before giving up. With this
system, if a backend does fail to clean a temporary schema due to
lock table exhaustion, autovacuum should hopefully put things right
the next time it processes the database.
Discussion: CAB7nPqSbYT6dRwsXVgiKmBdL_ARemfDZMPA+RPeC_ge0GK70hA@mail.gmail.com
Michael Paquier, with a bunch of comment changes by me.
We mustn't force parallel mode if the query has any subplans, since
ExecSerializePlan doesn't transmit them to workers. Testing
top_plan->initPlan is inadequate because (1) there might be initPlans
attached to lower plan nodes, and (2) non-initPlan subplans don't
work either. There's certainly room for improvement in those
restrictions, but for the moment that's what we've got.
Amit Kapila, per report from Andreas Seltenreich
Discussion: <8737im6pmh.fsf@credativ.de>
Because we use transformTargetList() for UPDATE as well as SELECT
tlists, the code accidentally tried to expand a "*" reference into
several columns. This is nonsensical, because the UPDATE syntax
provides exactly one target column to put the value into. The
immediate result was that transformUpdateTargetList() got confused
and reported "UPDATE target count mismatch --- internal error".
It seems better to treat such a reference as a plain whole-row
variable, as it would be in other contexts. (This could produce
useful results when the target column is of composite type.)
Fix by tweaking transformTargetList() to perform *-expansion only
conditionally, depending on its exprKind parameter.
Back-patch to 9.3. The problem exists further back, but a fix would be
much more invasive before that, because transformTargetList() wasn't
told what kind of list it was working on. Doesn't seem worth the
trouble given the lack of field reports. (I only noticed it because
I was checking the code while trying to improve the documentation about
how we handle "foo.*".)
Discussion: <4308.1479595330@sss.pgh.pa.us>
create_merge_append_path should use the path rowcount it just computed,
not rel->tuples, for costing purposes. Those numbers should always be
the same at present, but if we ever support parameterized MergeAppend
paths (a case this function is otherwise prepared for), the former would
be right and the latter wrong.
No need for back-patch since the problem is only latent.
Ashutosh Bapat
Discussion: <CAFjFpRek+cLCnTo24youuGtsq4zRphEB8EUUPjDxZjnL4n4HYQ@mail.gmail.com>
The serialization code dumped core for a string-valued GUC whose value
is NULL, which is a legal state. The infrastructure isn't capable of
transmitting that state exactly, but fortunately, transmitting an empty
string instead should be close enough (compare, eg, commit e45e990e4).
The code potentially underestimated the space required to format a
real-valued variable, both because it made an unwarranted assumption that
%g output would never be longer than %e output, and because it didn't count
right even for %e format. In practice this would pretty much always be
masked by overestimates for other variables, but it's still wrong.
Also fix boundary-case error in read_gucstate, incorrect handling of the
case where guc_sourcefile is non-NULL but zero length (not clear that can
happen, but if it did, this code would get totally confused), and
confusingly useless check for a NULL result from read_gucstate.
Andreas Seltenreich discovered the core dump; other issues noted while
reading nearby code. Back-patch to 9.5 where this code was introduced.
Michael Paquier and Tom Lane
Discussion: <871sy78wno.fsf@credativ.de>
Like pg_tables, pg_views, and others, this view contains information
about sequences in a way that is independent of the system catalog
layout but more comprehensive than the information schema.
To help implement the view, add a new internal function
pg_sequence_last_value() to return the last value of a sequence. This
is kept separate from pg_sequence_parameters() to separate querying
run-time state from catalog-like information.
Reviewed-by: Andreas Karlsson <andreas@proxel.se>
The debug messages that merely print StartTransactionCommand,
CommitTransactionCommand, ProcessUtilty, or ProcessQuery with no
additional details seem to be useless. Get rid of them.
The transaction status messages produced by ShowTransactionState are
occasionally useful, but they are extremely verbose, producing
multiple lines of log output every time they fire, which can happens
multiple times per transaction. So, reduce the level to DEBUG5; avoid
emitting an extra line just to explain which debug point is at issue;
and tighten up the rest of the message so it doesn't use quite so much
horizontal space.
With these changes, it's possible to run a somewhat busy system with a
log level even as high as DEBUG4, whereas previously anything above
DEBUG2 would flood the log with output that probably wasn't really all
that useful.
Replay of XLOG_BTREE_VACUUM during Hot Standby was previously thought to
require complex interlocking that matched the requirements on the
master. This required an O(N) operation that became a significant
problem with large indexes, causing replication delays of seconds or in
some cases minutes while the XLOG_BTREE_VACUUM was replayed.
This commit skips the “pin scan” that was previously required, by
observing in detail when and how it is safe to do so, with full
documentation. The pin scan is skipped only in replay; the VACUUM code
path on master is not touched here.
No tests included. Manual tests using an additional patch to view WAL records
and their timing have shown the change in WAL records and their handling has
successfully reduced replication delay.
This is a back-patch of commits 687f2cd7a0, 3e4b7d8798, b602842613
by Simon Riggs, to branches 9.4 and 9.5. No further backpatch is
possible because this depends on catalog scans being MVCC. I (Álvaro)
additionally updated a slight problem in the README, which explains why
this touches the 9.6 and master branches.
Previously, the handle for the control segment could not be zero, but
some other DSM segment could potentially have a handle value of zero.
However, that means that if someone wanted to store a dsm_handle that
might or might not be valid, they would need a separate boolean to
keep track of whether the associated value is legal. That's annoying,
so change things so that no DSM segment can ever have a handle of 0 -
or as we call it here, DSM_HANDLE_INVALID.
Thomas Munro. This was submitted as part of a much larger patch to
add an malloc-like allocator for dynamic shared memory, but this part
seems like a good idea independently of the rest of the patch.
The CatalogSnapshot was not plugged into SnapshotResetXmin()'s accounting
for whether MyPgXact->xmin could be cleared or advanced. In normal
transactions this was masked by the fact that the transaction snapshot
would be older, but during backend startup and certain utility commands
it was possible to re-use the CatalogSnapshot after MyPgXact->xmin had
been cleared, meaning that recently-deleted rows could be pruned even
though this snapshot could still see them, causing unexpected catalog
lookup failures. This effect appears to be the explanation for a recent
failure on buildfarm member piculet.
To fix, add the CatalogSnapshot to the RegisteredSnapshots heap whenever
it is valid.
In the previous logic, it was possible for the CatalogSnapshot to remain
valid across waits for client input, but with this change that would mean
it delays advance of global xmin in cases where it did not before. To
avoid possibly causing new table-bloat problems with clients that sit idle
for long intervals, add code to invalidate the CatalogSnapshot before
waiting for client input. (When the backend is busy, it's unlikely that
the CatalogSnapshot would be the oldest snap for very long, so we don't
worry about forcing early invalidation of it otherwise.)
In passing, remove the CatalogSnapshotStale flag in favor of using
"CatalogSnapshot != NULL" to represent validity, as we do for the other
special snapshots in snapmgr.c. And improve some obsolete comments.
No regression test because I don't know a deterministic way to cause this
failure. But the stress test shown in the original discussion provokes
"cache lookup failed for relation 1255" within a few dozen seconds for me.
Back-patch to 9.4 where MVCC catalog scans were introduced. (Note: it's
quite easy to produce similar failures with the same test case in branches
before 9.4. But MVCC catalog scans were supposed to fix that.)
Discussion: <16447.1478818294@sss.pgh.pa.us>
Be sure to pull up the subquery's hasRowSecurity flag when flattening a
subquery in pull_up_simple_subquery(). This isn't a bug today because
we don't look at the hasRowSecurity flag during planning, but it could
easily be a bug tomorrow.
Likewise, make rewriteRuleAction() pull up the hasRowSecurity flag when
absorbing RTEs from a rule action. This isn't a bug either, for the
opposite reason: the flag should never be set yet. But again, it seems
like good future proofing.
Add a comment explaining why rewriteTargetView() should *not* set
hasRowSecurity when adding stuff to securityQuals.
Improve some nearby comments about securityQuals processing, and document
that field more completely in parsenodes.h.
Patch by me, analysis by Dean Rasheed.
Discussion: <CAEZATCXZ8tb2DV6f=bkhsMV6u_gRcZ0CZBw2J-qU84RxSukZog@mail.gmail.com>
We just pass the data to the INSTEAD trigger.
Haribabu Kommi, reviewed by Dilip Kumar
Patch: <CAJrrPGcSQkrNkO+4PhLm4B8UQQQmU9YVUuqmtgM=pmzMfxWaWQ@mail.gmail.com>
Alvaro Herrera
Peter Eisentraut
Andres Freund
Robert Haas
Heikki Linnakangas
Tom Lane
Magnus Hagander