Make WAL segment size configurable at initdb time.

For performance reasons a larger segment size than the default 16MB
can be useful. A larger segment size has two main benefits: Firstly,
in setups using archiving, it makes it easier to write scripts that
can keep up with higher amounts of WAL, secondly, the WAL has to be
written and synced to disk less frequently.

But at the same time large segment size are disadvantageous for
smaller databases. So far the segment size had to be configured at
compile time, often making it unrealistic to choose one fitting to a
particularly load. Therefore change it to a initdb time setting.

This includes a breaking changes to the xlogreader.h API, which now
requires the current segment size to be configured.  For that and
similar reasons a number of binaries had to be taught how to recognize
the current segment size.

Author: Beena Emerson, editorialized by Andres Freund
Reviewed-By: Andres Freund, David Steele, Kuntal Ghosh, Michael
    Paquier, Peter Eisentraut, Robert Hass, Tushar Ahuja
Discussion: https://postgr.es/m/CAOG9ApEAcQ--1ieKbhFzXSQPw_YLmepaa4hNdnY5+ZULpt81Mw@mail.gmail.com

3 files changed, 31 insertions, 12 deletions

diff --git a/src/bin/pg_rewind/parsexlog.c b/src/bin/pg_rewind/parsexlog.c
index 1befdbdeea3..0fc71d2a135 100644
--- a/src/bin/pg_rewind/parsexlog.c
+++ b/src/bin/pg_rewind/parsexlog.c
@@ -69,7 +69,8 @@ extractPageMap(const char *datadir, XLogRecPtr startpoint, int tliIndex,
 
 	private.datadir = datadir;
 	private.tliIndex = tliIndex;
-	xlogreader = XLogReaderAllocate(&SimpleXLogPageRead, &private);
+	xlogreader = XLogReaderAllocate(WalSegSz, &SimpleXLogPageRead,
+									&private);
 	if (xlogreader == NULL)
 		pg_fatal("out of memory\n");
 
@@ -122,7 +123,8 @@ readOneRecord(const char *datadir, XLogRecPtr ptr, int tliIndex)
 
 	private.datadir = datadir;
 	private.tliIndex = tliIndex;
-	xlogreader = XLogReaderAllocate(&SimpleXLogPageRead, &private);
+	xlogreader = XLogReaderAllocate(WalSegSz, &SimpleXLogPageRead,
+									&private);
 	if (xlogreader == NULL)
 		pg_fatal("out of memory\n");
 
@@ -170,11 +172,17 @@ findLastCheckpoint(const char *datadir, XLogRecPtr forkptr, int tliIndex,
 	 * header in that case to find the next record.
 	 */
 	if (forkptr % XLOG_BLCKSZ == 0)
-		forkptr += (forkptr % XLogSegSize == 0) ? SizeOfXLogLongPHD : SizeOfXLogShortPHD;
+	{
+		if (XLogSegmentOffset(forkptr, WalSegSz) == 0)
+			forkptr += SizeOfXLogLongPHD;
+		else
+			forkptr += SizeOfXLogShortPHD;
+	}
 
 	private.datadir = datadir;
 	private.tliIndex = tliIndex;
-	xlogreader = XLogReaderAllocate(&SimpleXLogPageRead, &private);
+	xlogreader = XLogReaderAllocate(WalSegSz, &SimpleXLogPageRead,
+									&private);
 	if (xlogreader == NULL)
 		pg_fatal("out of memory\n");
 
@@ -239,21 +247,22 @@ SimpleXLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
 	XLogRecPtr	targetSegEnd;
 	XLogSegNo	targetSegNo;
 
-	XLByteToSeg(targetPagePtr, targetSegNo);
-	XLogSegNoOffsetToRecPtr(targetSegNo + 1, 0, targetSegEnd);
-	targetPageOff = targetPagePtr % XLogSegSize;
+	XLByteToSeg(targetPagePtr, targetSegNo, WalSegSz);
+	XLogSegNoOffsetToRecPtr(targetSegNo + 1, 0, targetSegEnd, WalSegSz);
+	targetPageOff = XLogSegmentOffset(targetPagePtr, WalSegSz);
 
 	/*
 	 * See if we need to switch to a new segment because the requested record
 	 * is not in the currently open one.
 	 */
-	if (xlogreadfd >= 0 && !XLByteInSeg(targetPagePtr, xlogreadsegno))
+	if (xlogreadfd >= 0 &&
+		!XLByteInSeg(targetPagePtr, xlogreadsegno, WalSegSz))
 	{
 		close(xlogreadfd);
 		xlogreadfd = -1;
 	}
 
-	XLByteToSeg(targetPagePtr, xlogreadsegno);
+	XLByteToSeg(targetPagePtr, xlogreadsegno, WalSegSz);
 
 	if (xlogreadfd < 0)
 	{
@@ -272,7 +281,8 @@ SimpleXLogPageRead(XLogReaderState *xlogreader, XLogRecPtr targetPagePtr,
 			   targetHistory[private->tliIndex].begin >= targetSegEnd)
 			private->tliIndex--;
 
-		XLogFileName(xlogfname, targetHistory[private->tliIndex].tli, xlogreadsegno);
+		XLogFileName(xlogfname, targetHistory[private->tliIndex].tli,
+					 xlogreadsegno, WalSegSz);
 
 		snprintf(xlogfpath, MAXPGPATH, "%s/" XLOGDIR "/%s", private->datadir, xlogfname);
 
diff --git a/src/bin/pg_rewind/pg_rewind.c b/src/bin/pg_rewind/pg_rewind.c
index 4bd1a759734..6079156e802 100644
--- a/src/bin/pg_rewind/pg_rewind.c
+++ b/src/bin/pg_rewind/pg_rewind.c
@@ -44,6 +44,7 @@ static ControlFileData ControlFile_target;
 static ControlFileData ControlFile_source;
 
 const char *progname;
+int			WalSegSz;
 
 /* Configuration options */
 char	   *datadir_target = NULL;
@@ -572,8 +573,8 @@ createBackupLabel(XLogRecPtr startpoint, TimeLineID starttli, XLogRecPtr checkpo
 	char		buf[1000];
 	int			len;
 
-	XLByteToSeg(startpoint, startsegno);
-	XLogFileName(xlogfilename, starttli, startsegno);
+	XLByteToSeg(startpoint, startsegno, WalSegSz);
+	XLogFileName(xlogfilename, starttli, startsegno, WalSegSz);
 
 	/*
 	 * Construct backup label file
@@ -631,6 +632,13 @@ digestControlFile(ControlFileData *ControlFile, char *src, size_t size)
 
 	memcpy(ControlFile, src, sizeof(ControlFileData));
 
+	/* set and validate WalSegSz */
+	WalSegSz = ControlFile->xlog_seg_size;
+
+	if (!IsValidWalSegSize(WalSegSz))
+		pg_fatal("WAL segment size must be a power of two between 1MB and 1GB, but the control file specifies %d bytes\n",
+				 WalSegSz);
+
 	/* Additional checks on control file */
 	checkControlFile(ControlFile);
 }
diff --git a/src/bin/pg_rewind/pg_rewind.h b/src/bin/pg_rewind/pg_rewind.h
index 31353dd3548..7bec34ff55d 100644
--- a/src/bin/pg_rewind/pg_rewind.h
+++ b/src/bin/pg_rewind/pg_rewind.h
@@ -24,6 +24,7 @@ extern char *connstr_source;
 extern bool debug;
 extern bool showprogress;
 extern bool dry_run;
+extern int	WalSegSz;
 
 /* Target history */
 extern TimeLineHistoryEntry *targetHistory;