Re: Lots of stuck queries after upgrade to 9.4

On 07/20/2015 03:01 PM, Andres Freund wrote:

Heikki,

On 2015-07-20 13:27:12 +0200, Andres Freund wrote:

On 2015-07-20 13:22:42 +0200, Andres Freund wrote:

Hm. The problem seems to be the WaitXLogInsertionsToFinish() call in
XLogFlush().

These are the relevant stack traces:
db9lock/debuglog-commit.txt
#2  0x00007f7405bd44f4 in LWLockWaitForVar (l=0x7f70f2ab6680, valptr=0x7f70f2ab66a0, oldval=<optimized out>, newval=0xffffffffffffffff) at /tmp/buildd/postgresql-9.4-9.4.4/build/../src/backend/storage/lmgr/lwlock.c:1011
#3  0x00007f7405a0d3e6 in WaitXLogInsertionsToFinish (upto=121713318915952) at /tmp/buildd/postgresql-9.4-9.4.4/build/../src/backend/access/transam/xlog.c:1755
#4  0x00007f7405a0e1d3 in XLogFlush (record=121713318911056) at /tmp/buildd/postgresql-9.4-9.4.4/build/../src/backend/access/transam/xlog.c:2849

db9lock/debuglog-insert-8276.txt
#1  0x00007f7405b77d91 in PGSemaphoreLock (sema=0x7f73ff6531d0, interruptOK=0 '\000') at pg_sema.c:421
#2  0x00007f7405bd4849 in LWLockAcquireCommon (val=<optimized out>, valptr=<optimized out>, mode=<optimized out>, l=<optimized out>) at /tmp/buildd/postgresql-9.4-9.4.4/build/../src/backend/storage/lmgr/lwlock.c:626
#3  LWLockAcquire (l=0x7f70ecaaa1a0, mode=LW_EXCLUSIVE) at /tmp/buildd/postgresql-9.4-9.4.4/build/../src/backend/storage/lmgr/lwlock.c:467
#4  0x00007f7405a0dcca in AdvanceXLInsertBuffer (upto=<optimized out>, opportunistic=<optimized out>) at /tmp/buildd/postgresql-9.4-9.4.4/build/../src/backend/access/transam/xlog.c:2161
#5  0x00007f7405a0e301 in GetXLogBuffer (ptr=121713318928384) at /tmp/buildd/postgresql-9.4-9.4.4/build/../src/backend/access/transam/xlog.c:1848
#6  0x00007f7405a0e9c9 in CopyXLogRecordToWAL (EndPos=<optimized out>, StartPos=<optimized out>, rdata=0x7ffff1c21b90, isLogSwitch=<optimized out>, write_len=<optimized out>) at /tmp/buildd/postgresql-9.4-9.4.4/build/../src/backend/access/transam/xlog.c:1494
#7  XLogInsert (rmid=<optimized out>, info=<optimized out>, rdata=<optimized out>) at /tmp/buildd/postgre

XLogFlush() has the following comment:
                        /*
                         * Re-check how far we can now flush the WAL. It's generally not
                         * safe to call WaitXLogInsertionsToFinish while holding
                         * WALWriteLock, because an in-progress insertion might need to
                         * also grab WALWriteLock to make progress. But we know that all
                         * the insertions up to insertpos have already finished, because
                         * that's what the earlier WaitXLogInsertionsToFinish() returned.
                         * We're only calling it again to allow insertpos to be moved
                         * further forward, not to actually wait for anyone.
                         */
                        insertpos = WaitXLogInsertionsToFinish(insertpos);

but I don't think that's valid reasoning. WaitXLogInsertionsToFinish()
calls LWLockWaitForVar(oldval = InvalidXLogRecPtr), which will block if
there's a exlusive locker and some backend doesn't yet have set
initializedUpto. Which seems like a ossible state?

A backend always updates its insert position before sleeping/acquiring another lock, by calling WALInsertLockUpdateInsertingAt. So even though another backend might indeed be in the initializedUpto==InvalidXlogRecPtr state, it will get out of that state before either by releasing the lock or updating initializedUpto, before it will in turn do anything that might deadlock.

Clearly there's *something* wrong here, though, given the bug report...

- Heikki