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

1 files changed, 22 insertions, 14 deletions

diff --git a/src/backend/access/transam/xlogutils.c b/src/backend/access/transam/xlogutils.c
index bbae733d658..b11c94c9b68 100644
--- a/src/backend/access/transam/xlogutils.c
+++ b/src/backend/access/transam/xlogutils.c
@@ -654,7 +654,8 @@ XLogTruncateRelation(RelFileNode rnode, ForkNumber forkNum,
  * frontend).  Probably these should be merged at some point.
  */
 static void
-XLogRead(char *buf, TimeLineID tli, XLogRecPtr startptr, Size count)
+XLogRead(char *buf, int segsize, TimeLineID tli, XLogRecPtr startptr,
+		 Size count)
 {
 	char	   *p;
 	XLogRecPtr	recptr;
@@ -666,6 +667,8 @@ XLogRead(char *buf, TimeLineID tli, XLogRecPtr startptr, Size count)
 	static TimeLineID sendTLI = 0;
 	static uint32 sendOff = 0;
 
+	Assert(segsize == wal_segment_size);
+
 	p = buf;
 	recptr = startptr;
 	nbytes = count;
@@ -676,10 +679,10 @@ XLogRead(char *buf, TimeLineID tli, XLogRecPtr startptr, Size count)
 		int			segbytes;
 		int			readbytes;
 
-		startoff = recptr % XLogSegSize;
+		startoff = XLogSegmentOffset(recptr, segsize);
 
 		/* Do we need to switch to a different xlog segment? */
-		if (sendFile < 0 || !XLByteInSeg(recptr, sendSegNo) ||
+		if (sendFile < 0 || !XLByteInSeg(recptr, sendSegNo, segsize) ||
 			sendTLI != tli)
 		{
 			char		path[MAXPGPATH];
@@ -687,9 +690,9 @@ XLogRead(char *buf, TimeLineID tli, XLogRecPtr startptr, Size count)
 			if (sendFile >= 0)
 				close(sendFile);
 
-			XLByteToSeg(recptr, sendSegNo);
+			XLByteToSeg(recptr, sendSegNo, segsize);
 
-			XLogFilePath(path, tli, sendSegNo);
+			XLogFilePath(path, tli, sendSegNo, segsize);
 
 			sendFile = BasicOpenFile(path, O_RDONLY | PG_BINARY, 0);
 
@@ -717,7 +720,7 @@ XLogRead(char *buf, TimeLineID tli, XLogRecPtr startptr, Size count)
 			{
 				char		path[MAXPGPATH];
 
-				XLogFilePath(path, tli, sendSegNo);
+				XLogFilePath(path, tli, sendSegNo, segsize);
 
 				ereport(ERROR,
 						(errcode_for_file_access(),
@@ -728,8 +731,8 @@ XLogRead(char *buf, TimeLineID tli, XLogRecPtr startptr, Size count)
 		}
 
 		/* How many bytes are within this segment? */
-		if (nbytes > (XLogSegSize - startoff))
-			segbytes = XLogSegSize - startoff;
+		if (nbytes > (segsize - startoff))
+			segbytes = segsize - startoff;
 		else
 			segbytes = nbytes;
 
@@ -740,7 +743,7 @@ XLogRead(char *buf, TimeLineID tli, XLogRecPtr startptr, Size count)
 		{
 			char		path[MAXPGPATH];
 
-			XLogFilePath(path, tli, sendSegNo);
+			XLogFilePath(path, tli, sendSegNo, segsize);
 
 			ereport(ERROR,
 					(errcode_for_file_access(),
@@ -798,7 +801,8 @@ XLogRead(char *buf, TimeLineID tli, XLogRecPtr startptr, Size count)
 void
 XLogReadDetermineTimeline(XLogReaderState *state, XLogRecPtr wantPage, uint32 wantLength)
 {
-	const XLogRecPtr lastReadPage = state->readSegNo * XLogSegSize + state->readOff;
+	const XLogRecPtr lastReadPage = state->readSegNo *
+		state->wal_segment_size + state->readOff;
 
 	Assert(wantPage != InvalidXLogRecPtr && wantPage % XLOG_BLCKSZ == 0);
 	Assert(wantLength <= XLOG_BLCKSZ);
@@ -842,7 +846,8 @@ XLogReadDetermineTimeline(XLogReaderState *state, XLogRecPtr wantPage, uint32 wa
 	if (state->currTLIValidUntil != InvalidXLogRecPtr &&
 		state->currTLI != ThisTimeLineID &&
 		state->currTLI != 0 &&
-		(wantPage + wantLength) / XLogSegSize < state->currTLIValidUntil / XLogSegSize)
+		((wantPage + wantLength) / state->wal_segment_size) <
+		(state->currTLIValidUntil / state->wal_segment_size))
 		return;
 
 	/*
@@ -864,9 +869,11 @@ XLogReadDetermineTimeline(XLogReaderState *state, XLogRecPtr wantPage, uint32 wa
 		 */
 		List	   *timelineHistory = readTimeLineHistory(ThisTimeLineID);
 
-		XLogRecPtr	endOfSegment = (((wantPage / XLogSegSize) + 1) * XLogSegSize) - 1;
+		XLogRecPtr	endOfSegment = (((wantPage / state->wal_segment_size) + 1)
+									* state->wal_segment_size) - 1;
 
-		Assert(wantPage / XLogSegSize == endOfSegment / XLogSegSize);
+		Assert(wantPage / state->wal_segment_size ==
+			   endOfSegment / state->wal_segment_size);
 
 		/*
 		 * Find the timeline of the last LSN on the segment containing
@@ -1014,7 +1021,8 @@ read_local_xlog_page(XLogReaderState *state, XLogRecPtr targetPagePtr,
 	 * as 'count', read the whole page anyway. It's guaranteed to be
 	 * zero-padded up to the page boundary if it's incomplete.
 	 */
-	XLogRead(cur_page, *pageTLI, targetPagePtr, XLOG_BLCKSZ);
+	XLogRead(cur_page, state->wal_segment_size, *pageTLI, targetPagePtr,
+			 XLOG_BLCKSZ);
 
 	/* number of valid bytes in the buffer */
 	return count;