database/postgres
database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-05-03 22:24:49 +03:00
Code

Repair sometimes-incorrect computation of StartUpID after a crash, per

Browse Source 
example from Rao Kumar.  This is a very corner corner-case, requiring
a minimum of three closely-spaced database crashes and an unlucky
positioning of the second recovery's checkpoint record before you'd notice
any problem.  But the consequences are dire enough that it's a must-fix.
This commit is contained in:
Tom Lane 2003-05-22 14:39:49 +00:00
parent 3284e342a2
commit bab030b288
1 changed files with 66 additions and 52 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

94
src/backend/access/transam/xlog.c
View File

@ -7,7 +7,7 @@
* Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
* Portions Copyright (c) 1994, Regents of the University of California
*
* $Header: /cvsroot/pgsql/src/backend/access/transam/xlog.c,v 1.109.2.1 2003/05/10 18:01:49 tgl Exp $
* $Header: /cvsroot/pgsql/src/backend/access/transam/xlog.c,v 1.109.2.2 2003/05/22 14:39:49 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -2559,7 +2559,18 @@ StartupXLOG(void)
ShmemVariableCache->nextOid = checkPoint.nextOid;
ShmemVariableCache->oidCount = 0;
/*
* If it was a shutdown checkpoint, then any following WAL entries were
* created under the next StartUpID; if it was a regular checkpoint then
* any following WAL entries were created under the same StartUpID.
* We must replay WAL entries using the same StartUpID they were created
* under, so temporarily adopt that SUI (see also xlog_redo()).
*/
if (wasShutdown)
ThisStartUpID = checkPoint.ThisStartUpID + 1;
else
ThisStartUpID = checkPoint.ThisStartUpID;
RedoRecPtr = XLogCtl->Insert.RedoRecPtr =
XLogCtl->SavedRedoRecPtr = checkPoint.redo;
@ -2654,39 +2665,6 @@ StartupXLOG(void)
ControlFile->logSeg = openLogSeg + 1;
Insert = &XLogCtl->Insert;
Insert->PrevRecord = LastRec;
/*
* If the next record will go to the new page then initialize for that
* one.
*/
if ((BLCKSZ - EndOfLog.xrecoff % BLCKSZ) < SizeOfXLogRecord)
EndOfLog.xrecoff += (BLCKSZ - EndOfLog.xrecoff % BLCKSZ);
if (EndOfLog.xrecoff % BLCKSZ == 0)
{
XLogRecPtr NewPageEndPtr;
NewPageEndPtr = EndOfLog;
if (NewPageEndPtr.xrecoff >= XLogFileSize)
{
/* crossing a logid boundary */
NewPageEndPtr.xlogid += 1;
NewPageEndPtr.xrecoff = BLCKSZ;
}
else
NewPageEndPtr.xrecoff += BLCKSZ;
XLogCtl->xlblocks[0] = NewPageEndPtr;
Insert->currpage->xlp_magic = XLOG_PAGE_MAGIC;
if (InRecovery)
Insert->currpage->xlp_sui = ThisStartUpID;
else
Insert->currpage->xlp_sui = ThisStartUpID + 1;
Insert->currpage->xlp_pageaddr.xlogid = NewPageEndPtr.xlogid;
Insert->currpage->xlp_pageaddr.xrecoff = NewPageEndPtr.xrecoff - BLCKSZ;
/* rest of buffer was zeroed in XLOGShmemInit */
Insert->currpos = (char *) Insert->currpage + SizeOfXLogPHD;
}
else
{
XLogCtl->xlblocks[0].xlogid = openLogId;
XLogCtl->xlblocks[0].xrecoff =
((EndOfLog.xrecoff - 1) / BLCKSZ + 1) * BLCKSZ;
@ -2701,8 +2679,7 @@ StartupXLOG(void)
Insert->currpos = (char *) Insert->currpage +
(EndOfLog.xrecoff + BLCKSZ - XLogCtl->xlblocks[0].xrecoff);
/* Make sure rest of page is zero */
memset(Insert->currpos, 0, INSERT_FREESPACE(Insert));
}
MemSet(Insert->currpos, 0, INSERT_FREESPACE(Insert));
LogwrtResult.Write = LogwrtResult.Flush = EndOfLog;
@ -2741,6 +2718,21 @@ StartupXLOG(void)
if (InRecovery)
{
/*
* At this point, ThisStartUpID is the largest SUI that we could
* find evidence for in the WAL entries. But check it against
* pg_control's latest checkpoint, to make sure that we can't
* accidentally re-use an already-used SUI.
*/
if (ThisStartUpID < ControlFile->checkPointCopy.ThisStartUpID)
ThisStartUpID = ControlFile->checkPointCopy.ThisStartUpID;
/*
* Perform a new checkpoint to update our recovery activity to disk.
*
* Note that we write a shutdown checkpoint. This is correct since
* the records following it will use SUI one more than what is shown
* in the checkpoint's ThisStartUpID.
*
* In case we had to use the secondary checkpoint, make sure that
* it will still be shown as the secondary checkpoint after this
* CreateCheckPoint operation; we don't want the broken primary
@ -2750,21 +2742,39 @@ StartupXLOG(void)
CreateCheckPoint(true, true);
XLogCloseRelationCache();
}
else
{
/*
* If we are not doing recovery, then we saw a checkpoint with nothing
* after it, and we can safely use StartUpID equal to one more than
* the checkpoint's SUI. But just for paranoia's sake, check against
* pg_control too.
*/
ThisStartUpID = checkPoint.ThisStartUpID;
if (ThisStartUpID < ControlFile->checkPointCopy.ThisStartUpID)
ThisStartUpID = ControlFile->checkPointCopy.ThisStartUpID;
}
/*
* Preallocate additional log files, if wanted.
*/
PreallocXlogFiles(EndOfLog);
/*
* Advance StartUpID to one more than the highest value used previously.
*/
ThisStartUpID++;
XLogCtl->ThisStartUpID = ThisStartUpID;
/*
* Okay, we're officially UP.
*/
InRecovery = false;
ControlFile->state = DB_IN_PRODUCTION;
ControlFile->time = time(NULL);
UpdateControlFile();
ThisStartUpID++;
XLogCtl->ThisStartUpID = ThisStartUpID;
/* Start up the commit log, too */
StartupCLOG();
@ -2960,7 +2970,7 @@ CreateCheckPoint(bool shutdown, bool force)
UpdateControlFile();
}
memset(&checkPoint, 0, sizeof(checkPoint));
MemSet(&checkPoint, 0, sizeof(checkPoint));
checkPoint.ThisStartUpID = ThisStartUpID;
checkPoint.time = time(NULL);
@ -3219,6 +3229,8 @@ xlog_redo(XLogRecPtr lsn, XLogRecord *record)
ShmemVariableCache->nextXid = checkPoint.nextXid;
ShmemVariableCache->nextOid = checkPoint.nextOid;
ShmemVariableCache->oidCount = 0;
/* Any later WAL records should be run with shutdown SUI plus 1 */
ThisStartUpID = checkPoint.ThisStartUpID + 1;
}
else if (info == XLOG_CHECKPOINT_ONLINE)
{
@ -3234,6 +3246,8 @@ xlog_redo(XLogRecPtr lsn, XLogRecord *record)
ShmemVariableCache->nextOid = checkPoint.nextOid;
ShmemVariableCache->oidCount = 0;
}
/* Any later WAL records should be run with the then-active SUI */
ThisStartUpID = checkPoint.ThisStartUpID;
}
}