Because this strategy logs changes on a block-by-block basis, it
avoids the need to checkpoint before and after the operation.
However, because it logs each changed block individually, it might
generate a lot of extra write-ahead logging if the template database
is large. Therefore, the older strategy remains available via a new
STRATEGY parameter to CREATE DATABASE, and a corresponding --strategy
option to createdb.
Somewhat controversially, this patch assembles the list of relations
to be copied to the new database by reading the pg_class relation of
the template database. Cross-database access like this isn't normally
possible, but it can be made to work here because there can't be any
connections to the database being copied, nor can it contain any
in-doubt transactions. Even so, we have to use lower-level interfaces
than normal, since the table scan and relcache interfaces will not
work for a database to which we're not connected. The advantage of
this approach is that we do not need to rely on the filesystem to
determine what ought to be copied, but instead on PostgreSQL's own
knowledge of the database structure. This avoids, for example,
copying stray files that happen to be located in the source database
directory.
Dilip Kumar, with a fairly large number of cosmetic changes by me.
Reviewed and tested by Ashutosh Sharma, Andres Freund, John Naylor,
Greg Nancarrow, Neha Sharma. Additional feedback from Bruce Momjian,
Heikki Linnakangas, Julien Rouhaud, Adam Brusselback, Kyotaro
Horiguchi, Tomas Vondra, Andrew Dunstan, Álvaro Herrera, and others.
Discussion: http://postgr.es/m/CA+TgmoYtcdxBjLh31DLxUXHxFVMPGzrU5_T=CYCvRyFHywSBUQ@mail.gmail.com
This commit improves the description of some WAL records for the
Transaction RMGR:
- Track remote_apply for a transaction commit. This GUC is
user-settable, so this information can be useful for debugging.
- Add replication origin information for PREPARE TRANSACTION, with the
origin ID, LSN and timestamp
- Same as above, for ROLLBACK PREPARED.
This impacts the format of pg_waldump or anything using these
description routines, so no backpatch is done.
Author: Masahiko Sawada, Michael Paquier
Discussion: https://postgr.es/m/CAD21AoD2dJfgsdxk4_KciAZMZQoUiCvmV9sDpp8ZuKLtKCNXaA@mail.gmail.com
Physical replication always ships WAL segment files to replicas once
they are complete. This is a problem if one WAL record is split across
a segment boundary and the primary server crashes before writing down
the segment with the next portion of the WAL record: WAL writing after
crash recovery would happily resume at the point where the broken record
started, overwriting that record ... but any standby or backup may have
already received a copy of that segment, and they are not rewinding.
This causes standbys to stop following the primary after the latter
crashes:
LOG: invalid contrecord length 7262 at A8/D9FFFBC8
because the standby is still trying to read the continuation record
(contrecord) for the original long WAL record, but it is not there and
it will never be. A workaround is to stop the replica, delete the WAL
file, and restart it -- at which point a fresh copy is brought over from
the primary. But that's pretty labor intensive, and I bet many users
would just give up and re-clone the standby instead.
A fix for this problem was already attempted in commit 515e3d84a0, but
it only addressed the case for the scenario of WAL archiving, so
streaming replication would still be a problem (as well as other things
such as taking a filesystem-level backup while the server is down after
having crashed), and it had performance scalability problems too; so it
had to be reverted.
This commit fixes the problem using an approach suggested by Andres
Freund, whereby the initial portion(s) of the split-up WAL record are
kept, and a special type of WAL record is written where the contrecord
was lost, so that WAL replay in the replica knows to skip the broken
parts. With this approach, we can continue to stream/archive segment
files as soon as they are complete, and replay of the broken records
will proceed across the crash point without a hitch.
Because a new type of WAL record is added, users should be careful to
upgrade standbys first, primaries later. Otherwise they risk the standby
being unable to start if the primary happens to write such a record.
A new TAP test that exercises this is added, but the portability of it
is yet to be seen.
This has been wrong since the introduction of physical replication, so
backpatch all the way back. In stable branches, keep the new
XLogReaderState members at the end of the struct, to avoid an ABI
break.
Author: Álvaro Herrera <alvherre@alvh.no-ip.org>
Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Reviewed-by: Nathan Bossart <bossartn@amazon.com>
Discussion: https://postgr.es/m/202108232252.dh7uxf6oxwcy@alvherre.pgsql
While tracking down the bug fixed in the preceding commit, I got quite
annoyed by the low quality of spg_desc's output. Add missing fields,
try to make the formatting consistent.
Commit 438fc4a39c prevented the WAL replay from writing
COMMIT_TS_SETTS record. By this change there is no code that
generates COMMIT_TS_SETTS record in PostgreSQL core.
Also we can think that there are no extensions using the record
because we've not received so far any complaints about the issue
that commit 438fc4a39c fixed. Therefore this commit removes
COMMIT_TS_SETTS record and its related code. Even without
this record, the timestamp required for commit timestamp feature
can be acquired from the COMMIT record.
Bump WAL page magic.
Reported-by: lx zou <zoulx1982@163.com>
Author: Fujii Masao
Reviewed-by: Alvaro Herrera
Discussion: https://postgr.es/m/16931-620d0f2fdc6108f1@postgresql.org
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
Remove the entire idea of "stale stats" within nbtree VACUUM (stop
caring about stats involving the number of inserted tuples). Also
remove the vacuum_cleanup_index_scale_factor GUC/param on the master
branch (though just disable them on postgres 13).
The vacuum_cleanup_index_scale_factor/stats interface made the nbtree AM
partially responsible for deciding when pg_class.reltuples stats needed
to be updated. This seems contrary to the spirit of the index AM API,
though -- it is not actually necessary for an index AM's bulk delete and
cleanup callbacks to provide accurate stats when it happens to be
inconvenient. The core code owns that. (Index AMs have the authority
to perform or not perform certain kinds of deferred cleanup based on
their own considerations, such as page deletion and recycling, but that
has little to do with pg_class.reltuples/num_index_tuples.)
This issue was fairly harmless until the introduction of the
autovacuum_vacuum_insert_threshold feature by commit b07642db, which had
an undesirable interaction with the vacuum_cleanup_index_scale_factor
mechanism: it made insert-driven autovacuums perform full index scans,
even though there is no real benefit to doing so. This has been tied to
a regression with an append-only insert benchmark [1].
Also have remaining cases that perform a full scan of an index during a
cleanup-only nbtree VACUUM indicate that the final tuple count is only
an estimate. This prevents vacuumlazy.c from setting the index's
pg_class.reltuples in those cases (it will now only update pg_class when
vacuumlazy.c had TIDs for nbtree to bulk delete). This arguably fixes
an oversight in deduplication-related bugfix commit 48e12913.
[1] https://smalldatum.blogspot.com/2021/01/insert-benchmark-postgres-is-still.html
Author: Peter Geoghegan <pg@bowt.ie>
Reviewed-By: Masahiko Sawada <sawada.mshk@gmail.com>
Discussion: https://postgr.es/m/CAD21AoA4WHthN5uU6+WScZ7+J_RcEjmcuH94qcoUPuB42ShXzg@mail.gmail.com
Backpatch: 13-, where autovacuum_vacuum_insert_threshold was added.
Otherwise we risk "leaking" deleted pages by making them non-recyclable
indefinitely. Commit 6655a729 did the same thing for deleted pages in
GiST indexes. That work was used as a starting point here.
Stop storing an XID indicating the oldest bpto.xact across all deleted
though unrecycled pages in nbtree metapages. There is no longer any
reason to care about that condition/the oldest XID. It only ever made
sense when wraparound was something _bt_vacuum_needs_cleanup() had to
consider.
The btm_oldest_btpo_xact metapage field has been repurposed and renamed.
It is now btm_last_cleanup_num_delpages, which is used to remember how
many non-recycled deleted pages remain from the last VACUUM (in practice
its value is usually the precise number of pages that were _newly
deleted_ during the specific VACUUM operation that last set the field).
The general idea behind storing btm_last_cleanup_num_delpages is to use
it to give _some_ consideration to non-recycled deleted pages inside
_bt_vacuum_needs_cleanup() -- though never too much. We only really
need to avoid leaving a truly excessive number of deleted pages in an
unrecycled state forever. We only do this to cover certain narrow cases
where no other factor makes VACUUM do a full scan, and yet the index
continues to grow (and so actually misses out on recycling existing
deleted pages).
These metapage changes result in a clear user-visible benefit: We no
longer trigger full index scans during VACUUM operations solely due to
the presence of only 1 or 2 known deleted (though unrecycled) blocks
from a very large index. All that matters now is keeping the costs and
benefits in balance over time.
Fix an issue that has been around since commit 857f9c36, which added the
"skip full scan of index" mechanism (i.e. the _bt_vacuum_needs_cleanup()
logic). The accuracy of btm_last_cleanup_num_heap_tuples accidentally
hinged upon _when_ the source value gets stored. We now always store
btm_last_cleanup_num_heap_tuples in btvacuumcleanup(). This fixes the
issue because IndexVacuumInfo.num_heap_tuples (the source field) is
expected to accurately indicate the state of the table _after_ the
VACUUM completes inside btvacuumcleanup().
A backpatchable fix cannot easily be extracted from this commit. A
targeted fix for the issue will follow in a later commit, though that
won't happen today.
I (pgeoghegan) have chosen to remove any mention of deleted pages in the
documentation of the vacuum_cleanup_index_scale_factor GUC/param, since
the presence of deleted (though unrecycled) pages is no longer of much
concern to users. The vacuum_cleanup_index_scale_factor description in
the docs now seems rather unclear in any case, and it should probably be
rewritten in the near future. Perhaps some passing mention of page
deletion will be added back at the same time.
Bump XLOG_PAGE_MAGIC due to nbtree WAL records using full XIDs now.
Author: Peter Geoghegan <pg@bowt.ie>
Reviewed-By: Masahiko Sawada <sawada.mshk@gmail.com>
Discussion: https://postgr.es/m/CAH2-WznpdHvujGUwYZ8sihX=d5u-tRYhi-F4wnV2uN2zHpMUXw@mail.gmail.com
Teach nbtree and heapam to cooperate in order to eagerly remove
duplicate tuples representing dead MVCC versions. This is "bottom-up
deletion". Each bottom-up deletion pass is triggered lazily in response
to a flood of versions on an nbtree leaf page. This usually involves a
"logically unchanged index" hint (these are produced by the executor
mechanism added by commit 9dc718bd).
The immediate goal of bottom-up index deletion is to avoid "unnecessary"
page splits caused entirely by version duplicates. It naturally has an
even more useful effect, though: it acts as a backstop against
accumulating an excessive number of index tuple versions for any given
_logical row_. Bottom-up index deletion complements what we might now
call "top-down index deletion": index vacuuming performed by VACUUM.
Bottom-up index deletion responds to the immediate local needs of
queries, while leaving it up to autovacuum to perform infrequent clean
sweeps of the index. The overall effect is to avoid certain
pathological performance issues related to "version churn" from UPDATEs.
The previous tableam interface used by index AMs to perform tuple
deletion (the table_compute_xid_horizon_for_tuples() function) has been
replaced with a new interface that supports certain new requirements.
Many (perhaps all) of the capabilities added to nbtree by this commit
could also be extended to other index AMs. That is left as work for a
later commit.
Extend deletion of LP_DEAD-marked index tuples in nbtree by adding logic
to consider extra index tuples (that are not LP_DEAD-marked) for
deletion in passing. This increases the number of index tuples deleted
significantly in many cases. The LP_DEAD deletion process (which is now
called "simple deletion" to clearly distinguish it from bottom-up
deletion) won't usually need to visit any extra table blocks to check
these extra tuples. We have to visit the same table blocks anyway to
generate a latestRemovedXid value (at least in the common case where the
index deletion operation's WAL record needs such a value).
Testing has shown that the "extra tuples" simple deletion enhancement
increases the number of index tuples deleted with almost any workload
that has LP_DEAD bits set in leaf pages. That is, it almost never fails
to delete at least a few extra index tuples. It helps most of all in
cases that happen to naturally have a lot of delete-safe tuples. It's
not uncommon for an individual deletion operation to end up deleting an
order of magnitude more index tuples compared to the old naive approach
(e.g., custom instrumentation of the patch shows that this happens
fairly often when the regression tests are run).
Add a further enhancement that augments simple deletion and bottom-up
deletion in indexes that make use of deduplication: Teach nbtree's
_bt_delitems_delete() function to support granular TID deletion in
posting list tuples. It is now possible to delete individual TIDs from
posting list tuples provided the TIDs have a tableam block number of a
table block that gets visited as part of the deletion process (visiting
the table block can be triggered directly or indirectly). Setting the
LP_DEAD bit of a posting list tuple is still an all-or-nothing thing,
but that matters much less now that deletion only needs to start out
with the right _general_ idea about which index tuples are deletable.
Bump XLOG_PAGE_MAGIC because xl_btree_delete changed.
No bump in BTREE_VERSION, since there are no changes to the on-disk
representation of nbtree indexes. Indexes built on PostgreSQL 12 or
PostgreSQL 13 will automatically benefit from bottom-up index deletion
(i.e. no reindexing required) following a pg_upgrade. The enhancement
to simple deletion is available with all B-Tree indexes following a
pg_upgrade, no matter what PostgreSQL version the user upgrades from.
Author: Peter Geoghegan <pg@bowt.ie>
Reviewed-By: Heikki Linnakangas <hlinnaka@iki.fi>
Reviewed-By: Victor Yegorov <vyegorov@gmail.com>
Discussion: https://postgr.es/m/CAH2-Wzm+maE3apHB8NOtmM=p-DO65j2V5GzAWCOEEuy3JZgb2g@mail.gmail.com
Bring heap and hash rmgr desc output in line with nbtree and GiST desc
output by using the name latestRemovedXid for all fields that output the
contents of the latestRemovedXid field from the WAL record's C struct
(stop using local variants).
This seems like a clear improvement because latestRemovedXid is a symbol
name that already appears across many different source files, and so is
probably much more recognizable.
Discussion: https://postgr.es/m/CAH2-Wzkt_Rs4VqPSCk87nyjPAAEmWL8STU9zgET_83EF5YfrLw@mail.gmail.com
A number of places were using appendStringInfo() when they could have been
using appendStringInfoString() instead. While there's no functionality
change there, it's just more efficient to use appendStringInfoString()
when no formatting is required. Likewise for some
appendStringInfoString() calls which were just appending a single char.
We can just use appendStringInfoChar() for that.
Additionally, many places were using appendPQExpBuffer() when they could
have used appendPQExpBufferStr(). Change those too.
Patch by Zhijie Hou, but further searching by me found significantly more
places that deserved the same treatment.
Author: Zhijie Hou, David Rowley
Discussion: https://postgr.es/m/cb172cf4361e4c7ba7167429070979d4@G08CNEXMBPEKD05.g08.fujitsu.local
Including Full in variable names duplicates the type information and
leads to overly long names. As FullTransactionId cannot accidentally
be casted to TransactionId that does not seem necessary.
Author: Andres Freund
Discussion: https://postgr.es/m/20200724011143.jccsyvsvymuiqfxu@alap3.anarazel.de
When wal_level=logical, write invalidations at command end into WAL so
that decoding can use this information.
This patch is required to allow the streaming of in-progress transactions
in logical decoding. The actual work to allow streaming will be committed
as a separate patch.
We still add the invalidations to the cache and write them to WAL at
commit time in RecordTransactionCommit(). This uses the existing
XLOG_INVALIDATIONS xlog record type, from the RM_STANDBY_ID resource
manager (see LogStandbyInvalidations for details).
So existing code relying on those invalidations (e.g. redo) does not need
to be changed.
The invalidations written at command end uses a new xlog record type
XLOG_XACT_INVALIDATIONS, from RM_XACT_ID resource manager. See
LogLogicalInvalidations for details.
These new xlog records are ignored by existing redo procedures, which
still rely on the invalidations written to commit records.
The invalidations are decoded and accumulated in top-transaction, and then
executed during replay. This obviates the need to decode the
invalidations as part of a commit record.
Bump XLOG_PAGE_MAGIC, since this introduces XLOG_XACT_INVALIDATIONS.
Author: Dilip Kumar, Tomas Vondra, Amit Kapila
Reviewed-by: Amit Kapila
Tested-by: Neha Sharma and Mahendra Singh Thalor
Discussion: https://postgr.es/m/688b0b7f-2f6c-d827-c27b-216a8e3ea700@2ndquadrant.com
Includes some manual cleanup of places that pgindent messed up,
most of which weren't per project style anyway.
Notably, it seems some people didn't absorb the style rules of
commit c9d297751, because there were a bunch of new occurrences
of function calls with a newline just after the left paren, all
with faulty expectations about how the rest of the call would get
indented.
An nbtree split point can be thought of as a point between two adjoining
tuples from an imaginary version of the page being split that includes
the incoming/new item (in addition to the items that really are on the
page). These adjoining tuples are called the lastleft and firstright
tuples.
The variables that represent split points contained a field called
firstright, which is an offset number of the first data item from the
original page that goes on the new right page. The corresponding tuple
from origpage was usually the same thing as the actual firstright tuple,
but not always: the firstright tuple is sometimes the new/incoming item
instead. This situation seems unnecessarily confusing.
Make things clearer by renaming the origpage offset returned by
_bt_findsplitloc() to "firstrightoff". We now have a firstright tuple
and a firstrightoff offset number which are comparable to the
newitem/lastleft tuples and the newitemoff/lastleftoff offset numbers
respectively. Also make sure that we are consistent about how we
describe nbtree page split point state.
Push the responsibility for dealing with pg_upgrade'd !heapkeyspace
indexes down to lower level code, relieving _bt_split() from dealing
with it directly. This means that we always have a palloc'd left page
high key on the leaf level, no matter what. This enables simplifying
some of the code (and code comments) within _bt_split().
Finally, restructure the page split code to make it clearer why suffix
truncation (which only takes place during leaf page splits) is
completely different to the first data item truncation that takes place
during internal page splits. Tuples are marked as having fewer
attributes stored in both cases, and the firstright tuple is truncated
in both cases, so it's easy to imagine somebody missing the distinction.
Until now, only selected bulk operations (e.g. COPY) did this. If a
given relfilenode received both a WAL-skipping COPY and a WAL-logged
operation (e.g. INSERT), recovery could lose tuples from the COPY. See
src/backend/access/transam/README section "Skipping WAL for New
RelFileNode" for the new coding rules. Maintainers of table access
methods should examine that section.
To maintain data durability, just before commit, we choose between an
fsync of the relfilenode and copying its contents to WAL. A new GUC,
wal_skip_threshold, guides that choice. If this change slows a workload
that creates small, permanent relfilenodes under wal_level=minimal, try
adjusting wal_skip_threshold. Users setting a timeout on COMMIT may
need to adjust that timeout, and log_min_duration_statement analysis
will reflect time consumption moving to COMMIT from commands like COPY.
Internally, this requires a reliable determination of whether
RollbackAndReleaseCurrentSubTransaction() would unlink a relation's
current relfilenode. Introduce rd_firstRelfilenodeSubid. Amend the
specification of rd_createSubid such that the field is zero when a new
rel has an old rd_node. Make relcache.c retain entries for certain
dropped relations until end of transaction.
Bump XLOG_PAGE_MAGIC, since this introduces XLOG_GIST_ASSIGN_LSN.
Future servers accept older WAL, so this bump is discretionary.
Kyotaro Horiguchi, reviewed (in earlier, similar versions) by Robert
Haas. Heikki Linnakangas and Michael Paquier implemented earlier
designs that materially clarified the problem. Reviewed, in earlier
designs, by Andrew Dunstan, Andres Freund, Alvaro Herrera, Tom Lane,
Fujii Masao, and Simon Riggs. Reported by Martijn van Oosterhout.
Discussion: https://postgr.es/m/20150702220524.GA9392@svana.org
Until now, only selected bulk operations (e.g. COPY) did this. If a
given relfilenode received both a WAL-skipping COPY and a WAL-logged
operation (e.g. INSERT), recovery could lose tuples from the COPY. See
src/backend/access/transam/README section "Skipping WAL for New
RelFileNode" for the new coding rules. Maintainers of table access
methods should examine that section.
To maintain data durability, just before commit, we choose between an
fsync of the relfilenode and copying its contents to WAL. A new GUC,
wal_skip_threshold, guides that choice. If this change slows a workload
that creates small, permanent relfilenodes under wal_level=minimal, try
adjusting wal_skip_threshold. Users setting a timeout on COMMIT may
need to adjust that timeout, and log_min_duration_statement analysis
will reflect time consumption moving to COMMIT from commands like COPY.
Internally, this requires a reliable determination of whether
RollbackAndReleaseCurrentSubTransaction() would unlink a relation's
current relfilenode. Introduce rd_firstRelfilenodeSubid. Amend the
specification of rd_createSubid such that the field is zero when a new
rel has an old rd_node. Make relcache.c retain entries for certain
dropped relations until end of transaction.
Back-patch to 9.5 (all supported versions). This introduces a new WAL
record type, XLOG_GIST_ASSIGN_LSN, without bumping XLOG_PAGE_MAGIC. As
always, update standby systems before master systems. This changes
sizeof(RelationData) and sizeof(IndexStmt), breaking binary
compatibility for affected extensions. (The most recent commit to
affect the same class of extensions was
089e4d405d0f3b94c74a2c6a54357a84a681754b.)
Kyotaro Horiguchi, reviewed (in earlier, similar versions) by Robert
Haas. Heikki Linnakangas and Michael Paquier implemented earlier
designs that materially clarified the problem. Reviewed, in earlier
designs, by Andrew Dunstan, Andres Freund, Alvaro Herrera, Tom Lane,
Fujii Masao, and Simon Riggs. Reported by Martijn van Oosterhout.
Discussion: https://postgr.es/m/20150702220524.GA9392@svana.org
Deduplication reduces the storage overhead of duplicates in indexes that
use the standard nbtree index access method. The deduplication process
is applied lazily, after the point where opportunistic deletion of
LP_DEAD-marked index tuples occurs. Deduplication is only applied at
the point where a leaf page split would otherwise be required. New
posting list tuples are formed by merging together existing duplicate
tuples. The physical representation of the items on an nbtree leaf page
is made more space efficient by deduplication, but the logical contents
of the page are not changed. Even unique indexes make use of
deduplication as a way of controlling bloat from duplicates whose TIDs
point to different versions of the same logical table row.
The lazy approach taken by nbtree has significant advantages over a GIN
style eager approach. Most individual inserts of index tuples have
exactly the same overhead as before. The extra overhead of
deduplication is amortized across insertions, just like the overhead of
page splits. The key space of indexes works in the same way as it has
since commit dd299df8 (the commit that made heap TID a tiebreaker
column).
Testing has shown that nbtree deduplication can generally make indexes
with about 10 or 15 tuples for each distinct key value about 2.5X - 4X
smaller, even with single column integer indexes (e.g., an index on a
referencing column that accompanies a foreign key). The final size of
single column nbtree indexes comes close to the final size of a similar
contrib/btree_gin index, at least in cases where GIN's posting list
compression isn't very effective. This can significantly improve
transaction throughput, and significantly reduce the cost of vacuuming
indexes.
A new index storage parameter (deduplicate_items) controls the use of
deduplication. The default setting is 'on', so all new B-Tree indexes
automatically use deduplication where possible. This decision will be
reviewed at the end of the Postgres 13 beta period.
There is a regression of approximately 2% of transaction throughput with
synthetic workloads that consist of append-only inserts into a table
with several non-unique indexes, where all indexes have few or no
repeated values. The underlying issue is that cycles are wasted on
unsuccessful attempts at deduplicating items in non-unique indexes.
There doesn't seem to be a way around it short of disabling
deduplication entirely. Note that deduplication of items in unique
indexes is fairly well targeted in general, which avoids the problem
there (we can use a special heuristic to trigger deduplication passes in
unique indexes, since we're specifically targeting "version bloat").
Bump XLOG_PAGE_MAGIC because xl_btree_vacuum changed.
No bump in BTREE_VERSION, since the representation of posting list
tuples works in a way that's backwards compatible with version 4 indexes
(i.e. indexes built on PostgreSQL 12). However, users must still
REINDEX a pg_upgrade'd index to use deduplication, regardless of the
Postgres version they've upgraded from. This is the only way to set the
new nbtree metapage flag indicating that deduplication is generally
safe.
Author: Anastasia Lubennikova, Peter Geoghegan
Reviewed-By: Peter Geoghegan, Heikki Linnakangas
Discussion:
https://postgr.es/m/55E4051B.7020209@postgrespro.ruhttps://postgr.es/m/4ab6e2db-bcee-f4cf-0916-3a06e6ccbb55@postgrespro.ru
We used to strategically place newlines after some function call left
parentheses to make pgindent move the argument list a few chars to the
left, so that the whole line would fit under 80 chars. However,
pgindent no longer does that, so the newlines just made the code
vertically longer for no reason. Remove those newlines, and reflow some
of those lines for some extra naturality.
Reviewed-by: Michael Paquier, Tom Lane
Discussion: https://postgr.es/m/20200129200401.GA6303@alvherre.pgsql
Commit 558a9165e0 taught _bt_delitems_delete() to produce its own XID
horizon on the primary. Standbys no longer needed to generate their own
latestRemovedXid, since they could just use the explicitly logged value
from the primary instead. The deleted offset numbers array from the
xl_btree_delete WAL record was no longer used by the REDO routine for
anything other than deleting the items.
This enables a minor optimization: We now treat the array as buffer
state, not generic WAL data, following _bt_delitems_vacuum()'s example.
This should be a minor win, since it allows us to avoid including the
deleted items array in cases where XLogInsert() stores the whole buffer
anyway. The primary goal here is to make the code more maintainable,
though. Removing inessential differences between the two functions
highlights the fundamental differences that remain.
Also change xl_btree_delete to use uint32 for the size of the array of
item offsets being deleted. This brings xl_btree_delete closer to
xl_btree_vacuum. Furthermore, it seems like a good idea to use an
explicit-width integer type (the field was previously an "int").
Bump XLOG_PAGE_MAGIC because xl_btree_delete changed.
Discussion: https://postgr.es/m/CAH2-Wzkz4TjmezzfAbaV1zYrh=fr0bCpzuJTvBe5iUQ3aUPsCQ@mail.gmail.com
The REDO routine for nbtree's xl_btree_vacuum record type hasn't
performed a "pin scan" since commit 3e4b7d87 went in, so clearly there
isn't any point in VACUUM WAL-logging information that won't actually be
used. Finish off the work of commit 3e4b7d87 (and the closely related
preceding commit 687f2cd7) by removing the code that generates this
unused information. Also remove the REDO routine code disabled by
commit 3e4b7d87.
Replace the unneeded lastBlockVacuumed field in xl_btree_vacuum with a
new "ndeleted" field. The new field isn't actually needed right now,
since we could continue to infer the array length from the overall
record length. However, an upcoming patch to add deduplication to
nbtree needs to add an "items updated" field to xl_btree_vacuum, so we
might as well start being explicit about the number of items now.
(Besides, it doesn't seem like a good idea to leave the xl_btree_vacuum
struct without any fields; the C standard says that that's undefined.)
nbtree VACUUM no longer forces writing a WAL record for the last block
in the index. Writing out a WAL record with no items for the final
block was supposed to force processing of a lastBlockVacuumed field by a
pin scan.
Bump XLOG_PAGE_MAGIC because xl_btree_vacuum changed.
Discussion: https://postgr.es/m/CAH2-WzmY_mT7UnTzFB5LBQDBkKpdV5UxP3B5bLb7uP%3D%3D6UQJRQ%40mail.gmail.com
Previously DROP DATABASE generated as many XLOG_DBASE_DROP WAL records
as the number of tablespaces that the database to drop uses. This caused
the scans of shared_buffers as many times as the number of the tablespaces
during recovery because WAL replay of one XLOG_DBASE_DROP record needs
that full scan. This could make the recovery time longer especially
when shared_buffers is large.
This commit changes DROP DATABASE so that it generates only one
XLOG_DBASE_DROP record, and registers the information of all the tablespaces
into it. Then, WAL replay of XLOG_DBASE_DROP record needs full scan of
shared_buffers only once, and which may improve the recovery performance.
Author: Fujii Masao
Reviewed-by: Kirk Jamison, Simon Riggs
Discussion: https://postgr.es/m/CAHGQGwF8YwNH0ZaL+2wjZPkj+ji9UhC+Z4ScnG97WKtVY5L9iw@mail.gmail.com
This commit changes xact_desc() so that it reports the detail information about
PREPARE TRANSACTION record, like GID (global transaction identifier),
timestamp at prepare transaction, delete-on-abort/commit relations,
XID of subtransactions, and invalidation messages. These are helpful
when diagnosing 2PC-related troubles.
Author: Fujii Masao
Reviewed-by: Michael Paquier, Andrey Lepikhov, Kyotaro Horiguchi, Julien Rouhaud, Alvaro Herrera
Discussion: https://postgr.es/m/CAHGQGwEvhASad4JJnCv=0dW2TJypZgW_Vpb-oZik2a3utCqcrA@mail.gmail.com
When maintaining or merging patches, one of the most common sources
for conflicts are the list of objects in makefiles. Especially when
the split across lines has been changed on both sides, which is
somewhat common due to attempting to stay below 80 columns, those
conflicts are unnecessarily laborious to resolve.
By splitting, and alphabetically sorting, OBJS style lines into one
object per line, conflicts should be less frequent, and easier to
resolve when they still occur.
Author: Andres Freund
Discussion: https://postgr.es/m/20191029200901.vww4idgcxv74cwes@alap3.anarazel.de
Include newitemoff in rmgr desc output for nbtree page split records.
In passing, correct an obsolete comment that claimed that newitemoff is
only logged for _L variant nbtree page split WAL records.
Both issues were oversights in commit 2c03216d83, which revamped the
WAL format.
Author: Peter Geoghegan
Backpatch: 9.5-, where the WAL format was revamped.
Otherwise, after a deleted page gets even older, it becomes unrecyclable
again. B-tree has the same problem, and has had since time immemorial,
but let's at least fix this in GiST, where this is new.
Backpatch to v12, where GiST page deletion was introduced.
Reviewed-by: Andrey Borodin
Discussion: https://www.postgresql.org/message-id/835A15A5-F1B4-4446-A711-BF48357EB602%40yandex-team.ru
This changes various places where appendPQExpBuffer was used in places
where it was possible to use appendPQExpBufferStr, and likewise for
appendStringInfo and appendStringInfoString. This is really just a
stylistic improvement, but there are also small performance gains to be
had from doing this.
Discussion: http://postgr.es/m/CAKJS1f9P=M-3ULmPvr8iCno8yvfDViHibJjpriHU8+SXUgeZ=w@mail.gmail.com
This record uses one metadata buffer and registers some data associated
to the buffer, but when parsing the record for its description a direct
access to the record data was done, but there is none. This leads
usually to an incorrect description, but can also cause crashes like in
pg_waldump. Instead, fix things so as the parsing uses the data
associated to the metadata block.
This is an oversight from 3d92796, so backpatch down to 11.
Author: Michael Paquier
Description: https://postgr.es/m/20190617013059.GA3153@paquier.xyz
Backpatch-through: 11
Due to parallel development, gist added the missing conflict
information in c952eae52a, while 558a9165e0 moved that computation
to the primary for the index types that already had it. Thus adapt
gist to also compute on the primary, using
index_compute_xid_horizon_for_tuples() instead of its own copy of the
logic.
This also adds pg_waldump support for XLOG_GIST_DELETE records, which
previously was not properly present.
Bumps WAL version.
Author: Andres Freund
Discussion: https://postgr.es/m/20190406050243.bszosdg4buvabfrt@alap3.anarazel.de
Instead of WAL-logging every modification during the build separately,
first build the index without any WAL-logging, and make a separate pass
through the index at the end, to write all pages to the WAL. This
significantly reduces the amount of WAL generated, and is usually also
faster, despite the extra I/O needed for the extra scan through the index.
WAL generated this way is also faster to replay.
For GiST, the LSN-NSN interlock makes this a little tricky. All pages must
be marked with a valid (i.e. non-zero) LSN, so that the parent-child
LSN-NSN interlock works correctly. We now use magic value 1 for that during
index build. Change the fake LSN counter to begin from 1000, so that 1 is
safely smaller than any real or fake LSN. 2 would've been enough for our
purposes, but let's reserve a bigger range, in case we need more special
values in the future.
Author: Anastasia Lubennikova, Andrey V. Lepikhov
Reviewed-by: Heikki Linnakangas, Dmitry Dolgov
Instead of inferring epoch progress from xids and checkpoints,
introduce a 64 bit FullTransactionId type and use it to track xid
generation. This fixes an unlikely bug where the epoch is reported
incorrectly if the range of active xids wraps around more than once
between checkpoints.
The only user-visible effect of this commit is to correct the epoch
used by txid_current() and txid_status(), also visible with
pg_controldata, in those rare circumstances. It also creates some
basic infrastructure so that later patches can use 64 bit
transaction IDs in more places.
The new type is a struct that we pass by value, as a form of strong
typedef. This prevents the sort of accidental confusion between
TransactionId and FullTransactionId that would be possible if we
were to use a plain old uint64.
Author: Thomas Munro
Reported-by: Amit Kapila
Reviewed-by: Andres Freund, Tom Lane, Heikki Linnakangas
Discussion: https://postgr.es/m/CAA4eK1%2BMv%2Bmb0HFfWM9Srtc6MVe160WFurXV68iAFMcagRZ0dQ%40mail.gmail.com
Previously the xid horizon was only computed during WAL replay. That
had two major problems:
1) It relied on knowing what the table pointed to looks like. That was
easy enough before the introducing of tableam (we knew it had to be
heap, although some trickery around logging the heap relfilenodes
was required). But to properly handle table AMs we need
per-database catalog access to look up the AM handler, which
recovery doesn't allow.
2) Not knowing the xid horizon also makes it hard to support logical
decoding on standbys. When on a catalog table, we need to be able
to conflict with slots that have an xid horizon that's too old. But
computing the horizon by visiting the heap only works once
consistency is reached, but we always need to be able to detect
conflicts.
There's also a secondary problem, in that the current method performs
redundant work on every standby. But that's counterbalanced by
potentially computing the value when not necessary (either because
there's no standby, or because there's no connected backends).
Solve 1) and 2) by moving computation of the xid horizon to the
primary and by involving tableam in the computation of the horizon.
To address the potentially increased overhead, increase the efficiency
of the xid horizon computation for heap by sorting the tids, and
eliminating redundant buffer accesses. When prefetching is available,
additionally perform prefetching of buffers. As this is more of a
maintenance task, rather than something routinely done in every read
only query, we add an arbitrary 10 to the effective concurrency -
thereby using IO concurrency, when not globally enabled. That's
possibly not the perfect formula, but seems good enough for now.
Bumps WAL format, as latestRemovedXid is now part of the records, and
the heap's relfilenode isn't anymore.
Author: Andres Freund, Amit Khandekar, Robert Haas
Reviewed-By: Robert Haas
Discussion:
https://postgr.es/m/20181212204154.nsxf3gzqv3gesl32@alap3.anarazel.dehttps://postgr.es/m/20181214014235.dal5ogljs3bmlq44@alap3.anarazel.dehttps://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
To do this, we scan GiST two times. In the first pass we make note of
empty leaf pages and internal pages. At second pass we scan through
internal pages, looking for downlinks to the empty pages.
Deleting internal pages is still not supported, like in nbtree, the last
child of an internal page is never deleted. That means that if you have a
workload where new keys are always inserted to different area than where
old keys are removed, the index will still grow without bound. But the rate
of growth will be an order of magnitude slower than before.
Author: Andrey Borodin
Discussion: https://www.postgresql.org/message-id/B1E4DF12-6CD3-4706-BDBD-BF3283328F60@yandex-team.ru
Make nbtree treat all index tuples as having a heap TID attribute.
Index searches can distinguish duplicates by heap TID, since heap TID is
always guaranteed to be unique. This general approach has numerous
benefits for performance, and is prerequisite to teaching VACUUM to
perform "retail index tuple deletion".
Naively adding a new attribute to every pivot tuple has unacceptable
overhead (it bloats internal pages), so suffix truncation of pivot
tuples is added. This will usually truncate away the "extra" heap TID
attribute from pivot tuples during a leaf page split, and may also
truncate away additional user attributes. This can increase fan-out,
especially in a multi-column index. Truncation can only occur at the
attribute granularity, which isn't particularly effective, but works
well enough for now. A future patch may add support for truncating
"within" text attributes by generating truncated key values using new
opclass infrastructure.
Only new indexes (BTREE_VERSION 4 indexes) will have insertions that
treat heap TID as a tiebreaker attribute, or will have pivot tuples
undergo suffix truncation during a leaf page split (on-disk
compatibility with versions 2 and 3 is preserved). Upgrades to version
4 cannot be performed on-the-fly, unlike upgrades from version 2 to
version 3. contrib/amcheck continues to work with version 2 and 3
indexes, while also enforcing stricter invariants when verifying version
4 indexes. These stricter invariants are the same invariants described
by "3.1.12 Sequencing" from the Lehman and Yao paper.
A later patch will enhance the logic used by nbtree to pick a split
point. This patch is likely to negatively impact performance without
smarter choices around the precise point to split leaf pages at. Making
these two mostly-distinct sets of enhancements into distinct commits
seems like it might clarify their design, even though neither commit is
particularly useful on its own.
The maximum allowed size of new tuples is reduced by an amount equal to
the space required to store an extra MAXALIGN()'d TID in a new high key
during leaf page splits. The user-facing definition of the "1/3 of a
page" restriction is already imprecise, and so does not need to be
revised. However, there should be a compatibility note in the v12
release notes.
Author: Peter Geoghegan
Reviewed-By: Heikki Linnakangas, Alexander Korotkov
Discussion: https://postgr.es/m/CAH2-WzkVb0Kom=R+88fDFb=JSxZMFvbHVC6Mn9LJ2n=X=kS-Uw@mail.gmail.com
Since its introduction, max_wal_senders is counted as part of
max_connections when it comes to define how many connection slots can be
used for replication connections with a WAL sender context. This can
lead to confusion for some users, as it could be possible to block a
base backup or replication from happening because other backend sessions
are already taken for other purposes by an application, and
superuser-only connection slots are not a correct solution to handle
that case.
This commit makes max_wal_senders independent of max_connections for its
handling of PGPROC entries in ProcGlobal, meaning that connection slots
for WAL senders are handled using their own free queue, like autovacuum
workers and bgworkers.
One compatibility issue that this change creates is that a standby now
requires to have a value of max_wal_senders at least equal to its
primary. So, if a standby created enforces the value of
max_wal_senders to be lower than that, then this could break failovers.
Normally this should not be an issue though, as any settings of a
standby are inherited from its primary as postgresql.conf gets normally
copied as part of a base backup, so parameters would be consistent.
Author: Alexander Kukushkin
Reviewed-by: Kyotaro Horiguchi, Petr Jelínek, Masahiko Sawada, Oleksii
Kliukin
Discussion: https://postgr.es/m/CAFh8B=nBzHQeYAu0b8fjK-AF1X4+_p6GRtwG+cCgs6Vci2uRuQ@mail.gmail.com
013ebc0a7b implements so-called GiST microvacuum. That is gistgettuple() marks
index tuples as dead when kill_prior_tuple is set. Later, when new tuple
insertion claims page space, those dead index tuples are physically deleted
from page. When this deletion is replayed on standby, it might conflict with
read-only queries. But 013ebc0a7b doesn't handle this. That may lead to
disappearance of some tuples from read-only snapshots on standby.
This commit implements resolving of conflicts between replay of GiST microvacuum
and standby queries. On the master we implement new WAL record type
XLOG_GIST_DELETE, which comprises necessary information. On stable releases
we've to be tricky to keep WAL compatibility. Information required for conflict
processing is just appended to data of XLOG_GIST_PAGE_UPDATE record. So,
PostgreSQL version, which doesn't know about conflict processing, will just
ignore that.
Reported-by: Andres Freund
Diagnosed-by: Andres Freund
Discussion: https://postgr.es/m/20181212224524.scafnlyjindmrbe6%40alap3.anarazel.de
Author: Alexander Korotkov
Backpatch-through: 9.6
Hexadecimal is consistently used as format to not bloat too much the
output but keep it readable. This information is useful mainly for
debugging purposes with for example pg_waldump.
Author: Michael Paquier
Reviewed-by: Nathan Bossart, Dmitry Dolgov, Andres Freund, Álvaro
Herrera
Discussion: https://postgr.es/m/20180413034734.GE1552@paquier.xyz
The "l" (ell) width spec means something in the corresponding scanf usage,
but not here. While modern POSIX says that applying "l" to "f" and other
floating format specs is a no-op, SUSv2 says it's undefined. Buildfarm
experience says that some old compilers emit warnings about it, and at
least one old stdio implementation (mingw's "ANSI" option) actually
produces wrong answers and/or crashes.
Discussion: https://postgr.es/m/21670.1526769114@sss.pgh.pa.us
Discussion: https://postgr.es/m/c085e1da-0d64-1c15-242d-c921f32e0d5c@dunslane.net
New WAL record XLOG_BTREE_META_CLEANUP introduced in 857f9c36 has no handling
in btree_desc() and btree_identify(). This patch implements corresponding
handling.
Alexander Korotkov
Review of commit 1eb6d652: It's pointless to add padding to the GID fields,
when the code that follows assumes that there is no alignment, and uses
memcpy(). Remove the pointless padding.
Update comments to note the new fields in the WAL records.
Reviewed-by: Michael Paquier
Discussion: https://www.postgresql.org/message-id/33b787bf-dc20-1161-54e9-3f3b607bf59d%40iki.fi
Robert Haas
Bruce Momjian
Michael Paquier
Alvaro Herrera
Tom Lane
Fujii Masao
Peter Geoghegan
Peter Eisentraut
David Rowley
Andres Freund
Amit Kapila
Noah Misch
Heikki Linnakangas
Thomas Munro
Alexander Korotkov
Peter Eisentraut
Teodor Sigaev