Prefetch data referenced by the WAL, take II.

Introduce a new GUC recovery_prefetch.  When enabled, look ahead in the
WAL and try to initiate asynchronous reading of referenced data blocks
that are not yet cached in our buffer pool.  For now, this is done with
posix_fadvise(), which has several caveats.  Since not all OSes have
that system call, "try" is provided so that it can be enabled where
available.  Better mechanisms for asynchronous I/O are possible in later
work.

Set to "try" for now for test coverage.  Default setting to be finalized
before release.

The GUC wal_decode_buffer_size limits the distance we can look ahead in
bytes of decoded data.

The existing GUC maintenance_io_concurrency is used to limit the number
of concurrent I/Os allowed, based on pessimistic heuristics used to
infer that I/Os have begun and completed.  We'll also not look more than
maintenance_io_concurrency * 4 block references ahead.

Reviewed-by: Julien Rouhaud <rjuju123@gmail.com>
Reviewed-by: Tomas Vondra <tomas.vondra@2ndquadrant.com>
Reviewed-by: Alvaro Herrera <alvherre@2ndquadrant.com> (earlier version)
Reviewed-by: Andres Freund <andres@anarazel.de> (earlier version)
Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> (earlier version)
Tested-by: Tomas Vondra <tomas.vondra@2ndquadrant.com> (earlier version)
Tested-by: Jakub Wartak <Jakub.Wartak@tomtom.com> (earlier version)
Tested-by: Dmitry Dolgov <9erthalion6@gmail.com> (earlier version)
Tested-by: Sait Talha Nisanci <Sait.Nisanci@microsoft.com> (earlier version)
Discussion: https://postgr.es/m/CA%2BhUKGJ4VJN8ttxScUFM8dOKX0BrBiboo5uz1cq%3DAovOddfHpA%40mail.gmail.com
This commit is contained in:
Thomas Munro 2022-04-07 19:28:40 +12:00
parent 9553b4115f
commit 5dc0418fab
27 changed files with 1595 additions and 77 deletions

View File

@ -3657,6 +3657,70 @@ include_dir 'conf.d'
</variablelist>
</sect2>
<sect2 id="runtime-config-wal-recovery">
<title>Recovery</title>
<indexterm>
<primary>configuration</primary>
<secondary>of recovery</secondary>
<tertiary>general settings</tertiary>
</indexterm>
<para>
This section describes the settings that apply to recovery in general,
affecting crash recovery, streaming replication and archive-based
replication.
</para>
<variablelist>
<varlistentry id="guc-recovery-prefetch" xreflabel="recovery_prefetch">
<term><varname>recovery_prefetch</varname> (<type>enum</type>)
<indexterm>
<primary><varname>recovery_prefetch</varname> configuration parameter</primary>
</indexterm>
</term>
<listitem>
<para>
Whether to try to prefetch blocks that are referenced in the WAL that
are not yet in the buffer pool, during recovery. Valid values are
<literal>off</literal> (the default), <literal>on</literal> and
<literal>try</literal>. The setting <literal>try</literal> enables
prefetching only if the operating system provides the
<function>posix_fadvise</function> function, which is currently used
to implement prefetching. Note that some operating systems provide the
function, but it doesn't do anything.
</para>
<para>
Prefetching blocks that will soon be needed can reduce I/O wait times
during recovery with some workloads.
See also the <xref linkend="guc-wal-decode-buffer-size"/> and
<xref linkend="guc-maintenance-io-concurrency"/> settings, which limit
prefetching activity.
</para>
</listitem>
</varlistentry>
<varlistentry id="guc-wal-decode-buffer-size" xreflabel="wal_decode_buffer_size">
<term><varname>wal_decode_buffer_size</varname> (<type>integer</type>)
<indexterm>
<primary><varname>wal_decode_buffer_size</varname> configuration parameter</primary>
</indexterm>
</term>
<listitem>
<para>
A limit on how far ahead the server can look in the WAL, to find
blocks to prefetch. If this value is specified without units, it is
taken as bytes.
The default is 512kB.
</para>
</listitem>
</varlistentry>
</variablelist>
</sect2>
<sect2 id="runtime-config-wal-archive-recovery">
<title>Archive Recovery</title>

View File

@ -328,6 +328,13 @@ postgres 27093 0.0 0.0 30096 2752 ? Ss 11:34 0:00 postgres: ser
</entry>
</row>
<row>
<entry><structname>pg_stat_recovery_prefetch</structname><indexterm><primary>pg_stat_recovery_prefetch</primary></indexterm></entry>
<entry>Only one row, showing statistics about blocks prefetched during recovery.
See <xref linkend="pg-stat-recovery-prefetch-view"/> for details.
</entry>
</row>
<row>
<entry><structname>pg_stat_subscription</structname><indexterm><primary>pg_stat_subscription</primary></indexterm></entry>
<entry>At least one row per subscription, showing information about
@ -2979,6 +2986,78 @@ SELECT pid, wait_event_type, wait_event FROM pg_stat_activity WHERE wait_event i
copy of the subscribed tables.
</para>
<table id="pg-stat-recovery-prefetch-view" xreflabel="pg_stat_recovery_prefetch">
<title><structname>pg_stat_recovery_prefetch</structname> View</title>
<tgroup cols="3">
<thead>
<row>
<entry>Column</entry>
<entry>Type</entry>
<entry>Description</entry>
</row>
</thead>
<tbody>
<row>
<entry><structfield>prefetch</structfield></entry>
<entry><type>bigint</type></entry>
<entry>Number of blocks prefetched because they were not in the buffer pool</entry>
</row>
<row>
<entry><structfield>hit</structfield></entry>
<entry><type>bigint</type></entry>
<entry>Number of blocks not prefetched because they were already in the buffer pool</entry>
</row>
<row>
<entry><structfield>skip_init</structfield></entry>
<entry><type>bigint</type></entry>
<entry>Number of blocks not prefetched because they would be zero-initialized</entry>
</row>
<row>
<entry><structfield>skip_new</structfield></entry>
<entry><type>bigint</type></entry>
<entry>Number of blocks not prefetched because they didn't exist yet</entry>
</row>
<row>
<entry><structfield>skip_fpw</structfield></entry>
<entry><type>bigint</type></entry>
<entry>Number of blocks not prefetched because a full page image was included in the WAL</entry>
</row>
<row>
<entry><structfield>skip_rep</structfield></entry>
<entry><type>bigint</type></entry>
<entry>Number of blocks not prefetched because they were already recently prefetched</entry>
</row>
<row>
<entry><structfield>wal_distance</structfield></entry>
<entry><type>integer</type></entry>
<entry>How many bytes ahead the prefetcher is looking</entry>
</row>
<row>
<entry><structfield>block_distance</structfield></entry>
<entry><type>integer</type></entry>
<entry>How many blocks ahead the prefetcher is looking</entry>
</row>
<row>
<entry><structfield>io_depth</structfield></entry>
<entry><type>integer</type></entry>
<entry>How many prefetches have been initiated but are not yet known to have completed</entry>
</row>
</tbody>
</tgroup>
</table>
<para>
The <structname>pg_stat_recovery_prefetch</structname> view will contain
only one row. It is filled with nulls if recovery has not run or
<xref linkend="guc-recovery-prefetch"/> is not enabled. The
columns <structfield>wal_distance</structfield>,
<structfield>block_distance</structfield>
and <structfield>io_depth</structfield> show current values, and the
other columns show cumulative counters that can be reset
with the <function>pg_stat_reset_shared</function> function.
</para>
<table id="pg-stat-subscription" xreflabel="pg_stat_subscription">
<title><structname>pg_stat_subscription</structname> View</title>
<tgroup cols="1">
@ -5199,8 +5278,11 @@ SELECT pid, wait_event_type, wait_event FROM pg_stat_activity WHERE wait_event i
all the counters shown in
the <structname>pg_stat_bgwriter</structname>
view, <literal>archiver</literal> to reset all the counters shown in
the <structname>pg_stat_archiver</structname> view or <literal>wal</literal>
to reset all the counters shown in the <structname>pg_stat_wal</structname> view.
the <structname>pg_stat_archiver</structname> view,
<literal>wal</literal> to reset all the counters shown in the
<structname>pg_stat_wal</structname> view or
<literal>recovery_prefetch</literal> to reset all the counters shown
in the <structname>pg_stat_recovery_prefetch</structname> view.
</para>
<para>
This function is restricted to superusers by default, but other users

View File

@ -803,6 +803,18 @@
counted as <literal>wal_write</literal> and <literal>wal_sync</literal>
in <structname>pg_stat_wal</structname>, respectively.
</para>
<para>
The <xref linkend="guc-recovery-prefetch"/> parameter can be used to reduce
I/O wait times during recovery by instructing the kernel to initiate reads
of disk blocks that will soon be needed but are not currently in
<productname>PostgreSQL</productname>'s buffer pool.
The <xref linkend="guc-maintenance-io-concurrency"/> and
<xref linkend="guc-wal-decode-buffer-size"/> settings limit prefetching
concurrency and distance, respectively. By default, it is set to
<literal>try</literal>, which enabled the feature on systems where
<function>posix_fadvise</function> is available.
</para>
</sect1>
<sect1 id="wal-internals">

View File

@ -31,6 +31,7 @@ OBJS = \
xlogarchive.o \
xlogfuncs.o \
xloginsert.o \
xlogprefetcher.o \
xlogreader.o \
xlogrecovery.o \
xlogutils.o

View File

@ -59,6 +59,7 @@
#include "access/xlog_internal.h"
#include "access/xlogarchive.h"
#include "access/xloginsert.h"
#include "access/xlogprefetcher.h"
#include "access/xlogreader.h"
#include "access/xlogrecovery.h"
#include "access/xlogutils.h"
@ -133,6 +134,7 @@ int CommitDelay = 0; /* precommit delay in microseconds */
int CommitSiblings = 5; /* # concurrent xacts needed to sleep */
int wal_retrieve_retry_interval = 5000;
int max_slot_wal_keep_size_mb = -1;
int wal_decode_buffer_size = 512 * 1024;
bool track_wal_io_timing = false;
#ifdef WAL_DEBUG

File diff suppressed because it is too large Load Diff

View File

@ -1727,6 +1727,8 @@ DecodeXLogRecord(XLogReaderState *state,
blk->has_image = ((fork_flags & BKPBLOCK_HAS_IMAGE) != 0);
blk->has_data = ((fork_flags & BKPBLOCK_HAS_DATA) != 0);
blk->prefetch_buffer = InvalidBuffer;
COPY_HEADER_FIELD(&blk->data_len, sizeof(uint16));
/* cross-check that the HAS_DATA flag is set iff data_length > 0 */
if (blk->has_data && blk->data_len == 0)
@ -1925,14 +1927,29 @@ err:
/*
* Returns information about the block that a block reference refers to.
*
* If the WAL record contains a block reference with the given ID, *rnode,
* *forknum, and *blknum are filled in (if not NULL), and returns true.
* Otherwise returns false.
* See XLogRecGetBlockTagExtended().
*/
bool
XLogRecGetBlockTag(XLogReaderState *record, uint8 block_id,
RelFileNode *rnode, ForkNumber *forknum, BlockNumber *blknum)
{
return XLogRecGetBlockTagExtended(record, block_id, rnode, forknum, blknum,
NULL);
}
/*
* Returns information about the block that a block reference refers to,
* optionally including the buffer that the block may already be in.
*
* If the WAL record contains a block reference with the given ID, *rnode,
* *forknum, *blknum and *prefetch_buffer are filled in (if not NULL), and
* returns true. Otherwise returns false.
*/
bool
XLogRecGetBlockTagExtended(XLogReaderState *record, uint8 block_id,
RelFileNode *rnode, ForkNumber *forknum,
BlockNumber *blknum,
Buffer *prefetch_buffer)
{
DecodedBkpBlock *bkpb;
@ -1947,6 +1964,8 @@ XLogRecGetBlockTag(XLogReaderState *record, uint8 block_id,
*forknum = bkpb->forknum;
if (blknum)
*blknum = bkpb->blkno;
if (prefetch_buffer)
*prefetch_buffer = bkpb->prefetch_buffer;
return true;
}

View File

@ -36,6 +36,7 @@
#include "access/xact.h"
#include "access/xlog_internal.h"
#include "access/xlogarchive.h"
#include "access/xlogprefetcher.h"
#include "access/xlogreader.h"
#include "access/xlogrecovery.h"
#include "access/xlogutils.h"
@ -183,6 +184,9 @@ static bool doRequestWalReceiverReply;
/* XLogReader object used to parse the WAL records */
static XLogReaderState *xlogreader = NULL;
/* XLogPrefetcher object used to consume WAL records with read-ahead */
static XLogPrefetcher *xlogprefetcher = NULL;
/* Parameters passed down from ReadRecord to the XLogPageRead callback. */
typedef struct XLogPageReadPrivate
{
@ -404,18 +408,21 @@ static void recoveryPausesHere(bool endOfRecovery);
static bool recoveryApplyDelay(XLogReaderState *record);
static void ConfirmRecoveryPaused(void);
static XLogRecord *ReadRecord(XLogReaderState *xlogreader,
int emode, bool fetching_ckpt, TimeLineID replayTLI);
static XLogRecord *ReadRecord(XLogPrefetcher *xlogprefetcher,
int emode, bool fetching_ckpt,
TimeLineID replayTLI);
static int XLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
int reqLen, XLogRecPtr targetRecPtr, char *readBuf);
static bool WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
bool fetching_ckpt,
XLogRecPtr tliRecPtr,
TimeLineID replayTLI,
XLogRecPtr replayLSN);
static XLogPageReadResult WaitForWALToBecomeAvailable(XLogRecPtr RecPtr,
bool randAccess,
bool fetching_ckpt,
XLogRecPtr tliRecPtr,
TimeLineID replayTLI,
XLogRecPtr replayLSN,
bool nonblocking);
static int emode_for_corrupt_record(int emode, XLogRecPtr RecPtr);
static XLogRecord *ReadCheckpointRecord(XLogReaderState *xlogreader, XLogRecPtr RecPtr,
static XLogRecord *ReadCheckpointRecord(XLogPrefetcher *xlogprefetcher, XLogRecPtr RecPtr,
int whichChkpt, bool report, TimeLineID replayTLI);
static bool rescanLatestTimeLine(TimeLineID replayTLI, XLogRecPtr replayLSN);
static int XLogFileRead(XLogSegNo segno, int emode, TimeLineID tli,
@ -561,6 +568,15 @@ InitWalRecovery(ControlFileData *ControlFile, bool *wasShutdown_ptr,
errdetail("Failed while allocating a WAL reading processor.")));
xlogreader->system_identifier = ControlFile->system_identifier;
/*
* Set the WAL decode buffer size. This limits how far ahead we can read
* in the WAL.
*/
XLogReaderSetDecodeBuffer(xlogreader, NULL, wal_decode_buffer_size);
/* Create a WAL prefetcher. */
xlogprefetcher = XLogPrefetcherAllocate(xlogreader);
/*
* Allocate two page buffers dedicated to WAL consistency checks. We do
* it this way, rather than just making static arrays, for two reasons:
@ -589,7 +605,8 @@ InitWalRecovery(ControlFileData *ControlFile, bool *wasShutdown_ptr,
* When a backup_label file is present, we want to roll forward from
* the checkpoint it identifies, rather than using pg_control.
*/
record = ReadCheckpointRecord(xlogreader, CheckPointLoc, 0, true, CheckPointTLI);
record = ReadCheckpointRecord(xlogprefetcher, CheckPointLoc, 0, true,
CheckPointTLI);
if (record != NULL)
{
memcpy(&checkPoint, XLogRecGetData(xlogreader), sizeof(CheckPoint));
@ -607,8 +624,8 @@ InitWalRecovery(ControlFileData *ControlFile, bool *wasShutdown_ptr,
*/
if (checkPoint.redo < CheckPointLoc)
{
XLogBeginRead(xlogreader, checkPoint.redo);
if (!ReadRecord(xlogreader, LOG, false,
XLogPrefetcherBeginRead(xlogprefetcher, checkPoint.redo);
if (!ReadRecord(xlogprefetcher, LOG, false,
checkPoint.ThisTimeLineID))
ereport(FATAL,
(errmsg("could not find redo location referenced by checkpoint record"),
@ -727,7 +744,7 @@ InitWalRecovery(ControlFileData *ControlFile, bool *wasShutdown_ptr,
CheckPointTLI = ControlFile->checkPointCopy.ThisTimeLineID;
RedoStartLSN = ControlFile->checkPointCopy.redo;
RedoStartTLI = ControlFile->checkPointCopy.ThisTimeLineID;
record = ReadCheckpointRecord(xlogreader, CheckPointLoc, 1, true,
record = ReadCheckpointRecord(xlogprefetcher, CheckPointLoc, 1, true,
CheckPointTLI);
if (record != NULL)
{
@ -1413,8 +1430,8 @@ FinishWalRecovery(void)
lastRec = XLogRecoveryCtl->lastReplayedReadRecPtr;
lastRecTLI = XLogRecoveryCtl->lastReplayedTLI;
}
XLogBeginRead(xlogreader, lastRec);
(void) ReadRecord(xlogreader, PANIC, false, lastRecTLI);
XLogPrefetcherBeginRead(xlogprefetcher, lastRec);
(void) ReadRecord(xlogprefetcher, PANIC, false, lastRecTLI);
endOfLog = xlogreader->EndRecPtr;
/*
@ -1503,6 +1520,9 @@ ShutdownWalRecovery(void)
{
char recoveryPath[MAXPGPATH];
/* Final update of pg_stat_recovery_prefetch. */
XLogPrefetcherComputeStats(xlogprefetcher);
/* Shut down xlogreader */
if (readFile >= 0)
{
@ -1510,6 +1530,7 @@ ShutdownWalRecovery(void)
readFile = -1;
}
XLogReaderFree(xlogreader);
XLogPrefetcherFree(xlogprefetcher);
if (ArchiveRecoveryRequested)
{
@ -1593,15 +1614,15 @@ PerformWalRecovery(void)
{
/* back up to find the record */
replayTLI = RedoStartTLI;
XLogBeginRead(xlogreader, RedoStartLSN);
record = ReadRecord(xlogreader, PANIC, false, replayTLI);
XLogPrefetcherBeginRead(xlogprefetcher, RedoStartLSN);
record = ReadRecord(xlogprefetcher, PANIC, false, replayTLI);
}
else
{
/* just have to read next record after CheckPoint */
Assert(xlogreader->ReadRecPtr == CheckPointLoc);
replayTLI = CheckPointTLI;
record = ReadRecord(xlogreader, LOG, false, replayTLI);
record = ReadRecord(xlogprefetcher, LOG, false, replayTLI);
}
if (record != NULL)
@ -1710,7 +1731,7 @@ PerformWalRecovery(void)
}
/* Else, try to fetch the next WAL record */
record = ReadRecord(xlogreader, LOG, false, replayTLI);
record = ReadRecord(xlogprefetcher, LOG, false, replayTLI);
} while (record != NULL);
/*
@ -1921,6 +1942,9 @@ ApplyWalRecord(XLogReaderState *xlogreader, XLogRecord *record, TimeLineID *repl
*/
if (AllowCascadeReplication())
WalSndWakeup();
/* Reset the prefetcher. */
XLogPrefetchReconfigure();
}
}
@ -2305,7 +2329,8 @@ verifyBackupPageConsistency(XLogReaderState *record)
* temporary page.
*/
buf = XLogReadBufferExtended(rnode, forknum, blkno,
RBM_NORMAL_NO_LOG);
RBM_NORMAL_NO_LOG,
InvalidBuffer);
if (!BufferIsValid(buf))
continue;
@ -2917,17 +2942,18 @@ ConfirmRecoveryPaused(void)
* Attempt to read the next XLOG record.
*
* Before first call, the reader needs to be positioned to the first record
* by calling XLogBeginRead().
* by calling XLogPrefetcherBeginRead().
*
* If no valid record is available, returns NULL, or fails if emode is PANIC.
* (emode must be either PANIC, LOG). In standby mode, retries until a valid
* record is available.
*/
static XLogRecord *
ReadRecord(XLogReaderState *xlogreader, int emode,
ReadRecord(XLogPrefetcher *xlogprefetcher, int emode,
bool fetching_ckpt, TimeLineID replayTLI)
{
XLogRecord *record;
XLogReaderState *xlogreader = XLogPrefetcherGetReader(xlogprefetcher);
XLogPageReadPrivate *private = (XLogPageReadPrivate *) xlogreader->private_data;
/* Pass through parameters to XLogPageRead */
@ -2943,7 +2969,7 @@ ReadRecord(XLogReaderState *xlogreader, int emode,
{
char *errormsg;
record = XLogReadRecord(xlogreader, &errormsg);
record = XLogPrefetcherReadRecord(xlogprefetcher, &errormsg);
if (record == NULL)
{
/*
@ -3056,9 +3082,12 @@ ReadRecord(XLogReaderState *xlogreader, int emode,
/*
* Read the XLOG page containing RecPtr into readBuf (if not read already).
* Returns number of bytes read, if the page is read successfully, or -1
* in case of errors. When errors occur, they are ereport'ed, but only
* if they have not been previously reported.
* Returns number of bytes read, if the page is read successfully, or
* XLREAD_FAIL in case of errors. When errors occur, they are ereport'ed, but
* only if they have not been previously reported.
*
* While prefetching, xlogreader->nonblocking may be set. In that case,
* returns XLREAD_WOULDBLOCK if we'd otherwise have to wait for more WAL.
*
* This is responsible for restoring files from archive as needed, as well
* as for waiting for the requested WAL record to arrive in standby mode.
@ -3066,7 +3095,7 @@ ReadRecord(XLogReaderState *xlogreader, int emode,
* 'emode' specifies the log level used for reporting "file not found" or
* "end of WAL" situations in archive recovery, or in standby mode when a
* trigger file is found. If set to WARNING or below, XLogPageRead() returns
* false in those situations, on higher log levels the ereport() won't
* XLREAD_FAIL in those situations, on higher log levels the ereport() won't
* return.
*
* In standby mode, if after a successful return of XLogPageRead() the
@ -3125,20 +3154,31 @@ retry:
(readSource == XLOG_FROM_STREAM &&
flushedUpto < targetPagePtr + reqLen))
{
if (!WaitForWALToBecomeAvailable(targetPagePtr + reqLen,
private->randAccess,
private->fetching_ckpt,
targetRecPtr,
private->replayTLI,
xlogreader->EndRecPtr))
{
if (readFile >= 0)
close(readFile);
readFile = -1;
readLen = 0;
readSource = XLOG_FROM_ANY;
if (readFile >= 0 &&
xlogreader->nonblocking &&
readSource == XLOG_FROM_STREAM &&
flushedUpto < targetPagePtr + reqLen)
return XLREAD_WOULDBLOCK;
return -1;
switch (WaitForWALToBecomeAvailable(targetPagePtr + reqLen,
private->randAccess,
private->fetching_ckpt,
targetRecPtr,
private->replayTLI,
xlogreader->EndRecPtr,
xlogreader->nonblocking))
{
case XLREAD_WOULDBLOCK:
return XLREAD_WOULDBLOCK;
case XLREAD_FAIL:
if (readFile >= 0)
close(readFile);
readFile = -1;
readLen = 0;
readSource = XLOG_FROM_ANY;
return XLREAD_FAIL;
case XLREAD_SUCCESS:
break;
}
}
@ -3263,7 +3303,7 @@ next_record_is_invalid:
if (StandbyMode)
goto retry;
else
return -1;
return XLREAD_FAIL;
}
/*
@ -3292,14 +3332,18 @@ next_record_is_invalid:
* available.
*
* When the requested record becomes available, the function opens the file
* containing it (if not open already), and returns true. When end of standby
* mode is triggered by the user, and there is no more WAL available, returns
* false.
* containing it (if not open already), and returns XLREAD_SUCCESS. When end
* of standby mode is triggered by the user, and there is no more WAL
* available, returns XLREAD_FAIL.
*
* If nonblocking is true, then give up immediately if we can't satisfy the
* request, returning XLREAD_WOULDBLOCK instead of waiting.
*/
static bool
static XLogPageReadResult
WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
bool fetching_ckpt, XLogRecPtr tliRecPtr,
TimeLineID replayTLI, XLogRecPtr replayLSN)
TimeLineID replayTLI, XLogRecPtr replayLSN,
bool nonblocking)
{
static TimestampTz last_fail_time = 0;
TimestampTz now;
@ -3353,6 +3397,14 @@ WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
*/
if (lastSourceFailed)
{
/*
* Don't allow any retry loops to occur during nonblocking
* readahead. Let the caller process everything that has been
* decoded already first.
*/
if (nonblocking)
return XLREAD_WOULDBLOCK;
switch (currentSource)
{
case XLOG_FROM_ARCHIVE:
@ -3367,7 +3419,7 @@ WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
if (StandbyMode && CheckForStandbyTrigger())
{
XLogShutdownWalRcv();
return false;
return XLREAD_FAIL;
}
/*
@ -3375,7 +3427,7 @@ WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
* and pg_wal.
*/
if (!StandbyMode)
return false;
return XLREAD_FAIL;
/*
* Move to XLOG_FROM_STREAM state, and set to start a
@ -3519,7 +3571,7 @@ WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
currentSource == XLOG_FROM_ARCHIVE ? XLOG_FROM_ANY :
currentSource);
if (readFile >= 0)
return true; /* success! */
return XLREAD_SUCCESS; /* success! */
/*
* Nope, not found in archive or pg_wal.
@ -3674,11 +3726,15 @@ WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
/* just make sure source info is correct... */
readSource = XLOG_FROM_STREAM;
XLogReceiptSource = XLOG_FROM_STREAM;
return true;
return XLREAD_SUCCESS;
}
break;
}
/* In nonblocking mode, return rather than sleeping. */
if (nonblocking)
return XLREAD_WOULDBLOCK;
/*
* Data not here yet. Check for trigger, then wait for
* walreceiver to wake us up when new WAL arrives.
@ -3686,13 +3742,13 @@ WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
if (CheckForStandbyTrigger())
{
/*
* Note that we don't "return false" immediately here.
* After being triggered, we still want to replay all
* the WAL that was already streamed. It's in pg_wal
* now, so we just treat this as a failure, and the
* state machine will move on to replay the streamed
* WAL from pg_wal, and then recheck the trigger and
* exit replay.
* Note that we don't return XLREAD_FAIL immediately
* here. After being triggered, we still want to
* replay all the WAL that was already streamed. It's
* in pg_wal now, so we just treat this as a failure,
* and the state machine will move on to replay the
* streamed WAL from pg_wal, and then recheck the
* trigger and exit replay.
*/
lastSourceFailed = true;
break;
@ -3711,6 +3767,9 @@ WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
streaming_reply_sent = true;
}
/* Update pg_stat_recovery_prefetch before sleeping. */
XLogPrefetcherComputeStats(xlogprefetcher);
/*
* Wait for more WAL to arrive. Time out after 5 seconds
* to react to a trigger file promptly and to check if the
@ -3743,7 +3802,7 @@ WaitForWALToBecomeAvailable(XLogRecPtr RecPtr, bool randAccess,
HandleStartupProcInterrupts();
}
return false; /* not reached */
return XLREAD_FAIL; /* not reached */
}
@ -3788,7 +3847,7 @@ emode_for_corrupt_record(int emode, XLogRecPtr RecPtr)
* 1 for "primary", 0 for "other" (backup_label)
*/
static XLogRecord *
ReadCheckpointRecord(XLogReaderState *xlogreader, XLogRecPtr RecPtr,
ReadCheckpointRecord(XLogPrefetcher *xlogprefetcher, XLogRecPtr RecPtr,
int whichChkpt, bool report, TimeLineID replayTLI)
{
XLogRecord *record;
@ -3815,8 +3874,8 @@ ReadCheckpointRecord(XLogReaderState *xlogreader, XLogRecPtr RecPtr,
return NULL;
}
XLogBeginRead(xlogreader, RecPtr);
record = ReadRecord(xlogreader, LOG, true, replayTLI);
XLogPrefetcherBeginRead(xlogprefetcher, RecPtr);
record = ReadRecord(xlogprefetcher, LOG, true, replayTLI);
if (record == NULL)
{

View File

@ -22,6 +22,7 @@
#include "access/timeline.h"
#include "access/xlogrecovery.h"
#include "access/xlog_internal.h"
#include "access/xlogprefetcher.h"
#include "access/xlogutils.h"
#include "miscadmin.h"
#include "pgstat.h"
@ -355,11 +356,13 @@ XLogReadBufferForRedoExtended(XLogReaderState *record,
RelFileNode rnode;
ForkNumber forknum;
BlockNumber blkno;
Buffer prefetch_buffer;
Page page;
bool zeromode;
bool willinit;
if (!XLogRecGetBlockTag(record, block_id, &rnode, &forknum, &blkno))
if (!XLogRecGetBlockTagExtended(record, block_id, &rnode, &forknum, &blkno,
&prefetch_buffer))
{
/* Caller specified a bogus block_id */
elog(PANIC, "failed to locate backup block with ID %d", block_id);
@ -381,7 +384,8 @@ XLogReadBufferForRedoExtended(XLogReaderState *record,
{
Assert(XLogRecHasBlockImage(record, block_id));
*buf = XLogReadBufferExtended(rnode, forknum, blkno,
get_cleanup_lock ? RBM_ZERO_AND_CLEANUP_LOCK : RBM_ZERO_AND_LOCK);
get_cleanup_lock ? RBM_ZERO_AND_CLEANUP_LOCK : RBM_ZERO_AND_LOCK,
prefetch_buffer);
page = BufferGetPage(*buf);
if (!RestoreBlockImage(record, block_id, page))
elog(ERROR, "failed to restore block image");
@ -410,7 +414,7 @@ XLogReadBufferForRedoExtended(XLogReaderState *record,
}
else
{
*buf = XLogReadBufferExtended(rnode, forknum, blkno, mode);
*buf = XLogReadBufferExtended(rnode, forknum, blkno, mode, prefetch_buffer);
if (BufferIsValid(*buf))
{
if (mode != RBM_ZERO_AND_LOCK && mode != RBM_ZERO_AND_CLEANUP_LOCK)
@ -450,6 +454,10 @@ XLogReadBufferForRedoExtended(XLogReaderState *record,
* exist, and we don't check for all-zeroes. Thus, no log entry is made
* to imply that the page should be dropped or truncated later.
*
* Optionally, recent_buffer can be used to provide a hint about the location
* of the page in the buffer pool; it does not have to be correct, but avoids
* a buffer mapping table probe if it is.
*
* NB: A redo function should normally not call this directly. To get a page
* to modify, use XLogReadBufferForRedoExtended instead. It is important that
* all pages modified by a WAL record are registered in the WAL records, or
@ -457,7 +465,8 @@ XLogReadBufferForRedoExtended(XLogReaderState *record,
*/
Buffer
XLogReadBufferExtended(RelFileNode rnode, ForkNumber forknum,
BlockNumber blkno, ReadBufferMode mode)
BlockNumber blkno, ReadBufferMode mode,
Buffer recent_buffer)
{
BlockNumber lastblock;
Buffer buffer;
@ -465,6 +474,15 @@ XLogReadBufferExtended(RelFileNode rnode, ForkNumber forknum,
Assert(blkno != P_NEW);
/* Do we have a clue where the buffer might be already? */
if (BufferIsValid(recent_buffer) &&
mode == RBM_NORMAL &&
ReadRecentBuffer(rnode, forknum, blkno, recent_buffer))
{
buffer = recent_buffer;
goto recent_buffer_fast_path;
}
/* Open the relation at smgr level */
smgr = smgropen(rnode, InvalidBackendId);
@ -523,6 +541,7 @@ XLogReadBufferExtended(RelFileNode rnode, ForkNumber forknum,
}
}
recent_buffer_fast_path:
if (mode == RBM_NORMAL)
{
/* check that page has been initialized */

View File

@ -930,6 +930,20 @@ CREATE VIEW pg_stat_wal_receiver AS
FROM pg_stat_get_wal_receiver() s
WHERE s.pid IS NOT NULL;
CREATE VIEW pg_stat_recovery_prefetch AS
SELECT
s.stats_reset,
s.prefetch,
s.hit,
s.skip_init,
s.skip_new,
s.skip_fpw,
s.skip_rep,
s.wal_distance,
s.block_distance,
s.io_depth
FROM pg_stat_get_recovery_prefetch() s;
CREATE VIEW pg_stat_subscription AS
SELECT
su.oid AS subid,

View File

@ -649,6 +649,8 @@ ReadRecentBuffer(RelFileNode rnode, ForkNumber forkNum, BlockNumber blockNum,
pg_atomic_write_u32(&bufHdr->state,
buf_state + BUF_USAGECOUNT_ONE);
pgBufferUsage.local_blks_hit++;
return true;
}
}
@ -680,6 +682,8 @@ ReadRecentBuffer(RelFileNode rnode, ForkNumber forkNum, BlockNumber blockNum,
else
PinBuffer_Locked(bufHdr); /* pin for first time */
pgBufferUsage.shared_blks_hit++;
return true;
}

View File

@ -211,7 +211,8 @@ XLogRecordPageWithFreeSpace(RelFileNode rnode, BlockNumber heapBlk,
blkno = fsm_logical_to_physical(addr);
/* If the page doesn't exist already, extend */
buf = XLogReadBufferExtended(rnode, FSM_FORKNUM, blkno, RBM_ZERO_ON_ERROR);
buf = XLogReadBufferExtended(rnode, FSM_FORKNUM, blkno, RBM_ZERO_ON_ERROR,
InvalidBuffer);
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
page = BufferGetPage(buf);

View File

@ -22,6 +22,7 @@
#include "access/subtrans.h"
#include "access/syncscan.h"
#include "access/twophase.h"
#include "access/xlogprefetcher.h"
#include "access/xlogrecovery.h"
#include "commands/async.h"
#include "miscadmin.h"
@ -119,6 +120,7 @@ CalculateShmemSize(int *num_semaphores)
size = add_size(size, LockShmemSize());
size = add_size(size, PredicateLockShmemSize());
size = add_size(size, ProcGlobalShmemSize());
size = add_size(size, XLogPrefetchShmemSize());
size = add_size(size, XLOGShmemSize());
size = add_size(size, XLogRecoveryShmemSize());
size = add_size(size, CLOGShmemSize());
@ -244,6 +246,7 @@ CreateSharedMemoryAndSemaphores(void)
* Set up xlog, clog, and buffers
*/
XLOGShmemInit();
XLogPrefetchShmemInit();
XLogRecoveryShmemInit();
CLOGShmemInit();
CommitTsShmemInit();

View File

@ -162,9 +162,11 @@ mdexists(SMgrRelation reln, ForkNumber forkNum)
{
/*
* Close it first, to ensure that we notice if the fork has been unlinked
* since we opened it.
* since we opened it. As an optimization, we can skip that in recovery,
* which already closes relations when dropping them.
*/
mdclose(reln, forkNum);
if (!InRecovery)
mdclose(reln, forkNum);
return (mdopenfork(reln, forkNum, EXTENSION_RETURN_NULL) != NULL);
}

View File

@ -16,6 +16,7 @@
#include "access/htup_details.h"
#include "access/xlog.h"
#include "access/xlogprefetcher.h"
#include "catalog/pg_authid.h"
#include "catalog/pg_type.h"
#include "common/ip.h"
@ -2103,13 +2104,15 @@ pg_stat_reset_shared(PG_FUNCTION_ARGS)
pgstat_reset_of_kind(PGSTAT_KIND_BGWRITER);
pgstat_reset_of_kind(PGSTAT_KIND_CHECKPOINTER);
}
else if (strcmp(target, "recovery_prefetch") == 0)
XLogPrefetchResetStats();
else if (strcmp(target, "wal") == 0)
pgstat_reset_of_kind(PGSTAT_KIND_WAL);
else
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("unrecognized reset target: \"%s\"", target),
errhint("Target must be \"archiver\", \"bgwriter\", or \"wal\".")));
errhint("Target must be \"archiver\", \"bgwriter\", \"recovery_prefetch\", or \"wal\".")));
PG_RETURN_VOID();
}

View File

@ -41,6 +41,7 @@
#include "access/twophase.h"
#include "access/xact.h"
#include "access/xlog_internal.h"
#include "access/xlogprefetcher.h"
#include "access/xlogrecovery.h"
#include "catalog/namespace.h"
#include "catalog/objectaccess.h"
@ -217,6 +218,7 @@ static bool check_effective_io_concurrency(int *newval, void **extra, GucSource
static bool check_maintenance_io_concurrency(int *newval, void **extra, GucSource source);
static bool check_huge_page_size(int *newval, void **extra, GucSource source);
static bool check_client_connection_check_interval(int *newval, void **extra, GucSource source);
static void assign_maintenance_io_concurrency(int newval, void *extra);
static bool check_application_name(char **newval, void **extra, GucSource source);
static void assign_application_name(const char *newval, void *extra);
static bool check_cluster_name(char **newval, void **extra, GucSource source);
@ -495,6 +497,19 @@ static const struct config_enum_entry huge_pages_options[] = {
{NULL, 0, false}
};
static const struct config_enum_entry recovery_prefetch_options[] = {
{"off", RECOVERY_PREFETCH_OFF, false},
{"on", RECOVERY_PREFETCH_ON, false},
{"try", RECOVERY_PREFETCH_TRY, false},
{"true", RECOVERY_PREFETCH_ON, true},
{"false", RECOVERY_PREFETCH_OFF, true},
{"yes", RECOVERY_PREFETCH_ON, true},
{"no", RECOVERY_PREFETCH_OFF, true},
{"1", RECOVERY_PREFETCH_ON, true},
{"0", RECOVERY_PREFETCH_OFF, true},
{NULL, 0, false}
};
static const struct config_enum_entry force_parallel_mode_options[] = {
{"off", FORCE_PARALLEL_OFF, false},
{"on", FORCE_PARALLEL_ON, false},
@ -785,6 +800,8 @@ const char *const config_group_names[] =
gettext_noop("Write-Ahead Log / Checkpoints"),
/* WAL_ARCHIVING */
gettext_noop("Write-Ahead Log / Archiving"),
/* WAL_RECOVERY */
gettext_noop("Write-Ahead Log / Recovery"),
/* WAL_ARCHIVE_RECOVERY */
gettext_noop("Write-Ahead Log / Archive Recovery"),
/* WAL_RECOVERY_TARGET */
@ -2818,6 +2835,17 @@ static struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
{
{"wal_decode_buffer_size", PGC_POSTMASTER, WAL_RECOVERY,
gettext_noop("Maximum buffer size for reading ahead in the WAL during recovery."),
gettext_noop("This controls the maximum distance we can read ahead in the WAL to prefetch referenced blocks."),
GUC_UNIT_BYTE
},
&wal_decode_buffer_size,
512 * 1024, 64 * 1024, MaxAllocSize,
NULL, NULL, NULL
},
{
{"wal_keep_size", PGC_SIGHUP, REPLICATION_SENDING,
gettext_noop("Sets the size of WAL files held for standby servers."),
@ -3141,7 +3169,8 @@ static struct config_int ConfigureNamesInt[] =
0,
#endif
0, MAX_IO_CONCURRENCY,
check_maintenance_io_concurrency, NULL, NULL
check_maintenance_io_concurrency, assign_maintenance_io_concurrency,
NULL
},
{
@ -5013,6 +5042,16 @@ static struct config_enum ConfigureNamesEnum[] =
NULL, NULL, NULL
},
{
{"recovery_prefetch", PGC_SIGHUP, WAL_RECOVERY,
gettext_noop("Prefetch referenced blocks during recovery"),
gettext_noop("Look ahead in the WAL to find references to uncached data.")
},
&recovery_prefetch,
RECOVERY_PREFETCH_TRY, recovery_prefetch_options,
check_recovery_prefetch, assign_recovery_prefetch, NULL
},
{
{"force_parallel_mode", PGC_USERSET, DEVELOPER_OPTIONS,
gettext_noop("Forces use of parallel query facilities."),
@ -12422,6 +12461,20 @@ check_client_connection_check_interval(int *newval, void **extra, GucSource sour
return true;
}
static void
assign_maintenance_io_concurrency(int newval, void *extra)
{
#ifdef USE_PREFETCH
/*
* Reconfigure recovery prefetching, because a setting it depends on
* changed.
*/
maintenance_io_concurrency = newval;
if (AmStartupProcess())
XLogPrefetchReconfigure();
#endif
}
static bool
check_application_name(char **newval, void **extra, GucSource source)
{

View File

@ -241,6 +241,12 @@
#max_wal_size = 1GB
#min_wal_size = 80MB
# - Prefetching during recovery -
#recovery_prefetch = try # prefetch pages referenced in the WAL?
#wal_decode_buffer_size = 512kB # lookahead window used for prefetching
# (change requires restart)
# - Archiving -
#archive_mode = off # enables archiving; off, on, or always

View File

@ -50,6 +50,7 @@ extern bool *wal_consistency_checking;
extern char *wal_consistency_checking_string;
extern bool log_checkpoints;
extern bool track_wal_io_timing;
extern int wal_decode_buffer_size;
extern int CheckPointSegments;

View File

@ -0,0 +1,53 @@
/*-------------------------------------------------------------------------
*
* xlogprefetcher.h
* Declarations for the recovery prefetching module.
*
* Portions Copyright (c) 2022, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
* src/include/access/xlogprefetcher.h
*-------------------------------------------------------------------------
*/
#ifndef XLOGPREFETCHER_H
#define XLOGPREFETCHER_H
#include "access/xlogdefs.h"
/* GUCs */
extern int recovery_prefetch;
/* Possible values for recovery_prefetch */
typedef enum
{
RECOVERY_PREFETCH_OFF,
RECOVERY_PREFETCH_ON,
RECOVERY_PREFETCH_TRY
} RecoveryPrefetchValue;
struct XLogPrefetcher;
typedef struct XLogPrefetcher XLogPrefetcher;
extern void XLogPrefetchReconfigure(void);
extern size_t XLogPrefetchShmemSize(void);
extern void XLogPrefetchShmemInit(void);
extern void XLogPrefetchResetStats(void);
extern XLogPrefetcher *XLogPrefetcherAllocate(XLogReaderState *reader);
extern void XLogPrefetcherFree(XLogPrefetcher *prefetcher);
extern XLogReaderState *XLogPrefetcherGetReader(XLogPrefetcher *prefetcher);
extern void XLogPrefetcherBeginRead(XLogPrefetcher *prefetcher,
XLogRecPtr recPtr);
extern XLogRecord *XLogPrefetcherReadRecord(XLogPrefetcher *prefetcher,
char **errmsg);
extern void XLogPrefetcherComputeStats(XLogPrefetcher *prefetcher);
#endif

View File

@ -39,6 +39,7 @@
#endif
#include "access/xlogrecord.h"
#include "storage/buf.h"
/* WALOpenSegment represents a WAL segment being read. */
typedef struct WALOpenSegment
@ -125,6 +126,9 @@ typedef struct
ForkNumber forknum;
BlockNumber blkno;
/* Prefetching workspace. */
Buffer prefetch_buffer;
/* copy of the fork_flags field from the XLogRecordBlockHeader */
uint8 flags;
@ -430,5 +434,9 @@ extern char *XLogRecGetBlockData(XLogReaderState *record, uint8 block_id, Size *
extern bool XLogRecGetBlockTag(XLogReaderState *record, uint8 block_id,
RelFileNode *rnode, ForkNumber *forknum,
BlockNumber *blknum);
extern bool XLogRecGetBlockTagExtended(XLogReaderState *record, uint8 block_id,
RelFileNode *rnode, ForkNumber *forknum,
BlockNumber *blknum,
Buffer *prefetch_buffer);
#endif /* XLOGREADER_H */

View File

@ -84,7 +84,8 @@ extern XLogRedoAction XLogReadBufferForRedoExtended(XLogReaderState *record,
Buffer *buf);
extern Buffer XLogReadBufferExtended(RelFileNode rnode, ForkNumber forknum,
BlockNumber blkno, ReadBufferMode mode);
BlockNumber blkno, ReadBufferMode mode,
Buffer recent_buffer);
extern Relation CreateFakeRelcacheEntry(RelFileNode rnode);
extern void FreeFakeRelcacheEntry(Relation fakerel);

View File

@ -53,6 +53,6 @@
*/
/* yyyymmddN */
#define CATALOG_VERSION_NO 202204073
#define CATALOG_VERSION_NO 202204074
#endif

View File

@ -5654,6 +5654,13 @@
proargmodes => '{o,o,o,o,o,o,o,o,o}',
proargnames => '{wal_records,wal_fpi,wal_bytes,wal_buffers_full,wal_write,wal_sync,wal_write_time,wal_sync_time,stats_reset}',
prosrc => 'pg_stat_get_wal' },
{ oid => '9085', descr => 'statistics: information about WAL prefetching',
proname => 'pg_stat_get_recovery_prefetch', prorows => '1', provolatile => 'v',
proretset => 't', prorettype => 'record', proargtypes => '',
proallargtypes => '{timestamptz,int8,int8,int8,int8,int8,int8,int4,int4,int4}',
proargmodes => '{o,o,o,o,o,o,o,o,o,o}',
proargnames => '{stats_reset,prefetch,hit,skip_init,skip_new,skip_fpw,skip_rep,wal_distance,block_distance,io_depth}',
prosrc => 'pg_stat_get_recovery_prefetch' },
{ oid => '2306', descr => 'statistics: information about SLRU caches',
proname => 'pg_stat_get_slru', prorows => '100', proisstrict => 'f',

View File

@ -453,4 +453,8 @@ extern void assign_search_path(const char *newval, void *extra);
extern bool check_wal_buffers(int *newval, void **extra, GucSource source);
extern void assign_xlog_sync_method(int new_sync_method, void *extra);
/* in access/transam/xlogprefetcher.c */
extern bool check_recovery_prefetch(int *new_value, void **extra, GucSource source);
extern void assign_recovery_prefetch(int new_value, void *extra);
#endif /* GUC_H */

View File

@ -67,6 +67,7 @@ enum config_group
WAL_SETTINGS,
WAL_CHECKPOINTS,
WAL_ARCHIVING,
WAL_RECOVERY,
WAL_ARCHIVE_RECOVERY,
WAL_RECOVERY_TARGET,
REPLICATION_SENDING,

View File

@ -2019,6 +2019,17 @@ pg_stat_progress_vacuum| SELECT s.pid,
s.param7 AS num_dead_tuples
FROM (pg_stat_get_progress_info('VACUUM'::text) s(pid, datid, relid, param1, param2, param3, param4, param5, param6, param7, param8, param9, param10, param11, param12, param13, param14, param15, param16, param17, param18, param19, param20)
LEFT JOIN pg_database d ON ((s.datid = d.oid)));
pg_stat_recovery_prefetch| SELECT s.stats_reset,
s.prefetch,
s.hit,
s.skip_init,
s.skip_new,
s.skip_fpw,
s.skip_rep,
s.wal_distance,
s.block_distance,
s.io_depth
FROM pg_stat_get_recovery_prefetch() s(stats_reset, prefetch, hit, skip_init, skip_new, skip_fpw, skip_rep, wal_distance, block_distance, io_depth);
pg_stat_replication| SELECT s.pid,
s.usesysid,
u.rolname AS usename,

View File

@ -1421,6 +1421,9 @@ LogicalRepWorker
LogicalRewriteMappingData
LogicalTape
LogicalTapeSet
LsnReadQueue
LsnReadQueueNextFun
LsnReadQueueNextStatus
LtreeGistOptions
LtreeSignature
MAGIC
@ -2949,6 +2952,9 @@ XLogPageHeaderData
XLogPageReadCB
XLogPageReadPrivate
XLogPageReadResult
XLogPrefetcher
XLogPrefetcherFilter
XLogPrefetchStats
XLogReaderRoutine
XLogReaderState
XLogRecData