None of the locks are in state false actually. I don't have iostat on that machine, but vmstat shows a lot of writes to the drives, and the runnable processes are more than 1:
procs memory swap io system cpu r b swpd free buff cache si so bi bo in cs us sy wa id 1 2 0 3857568 292936 2791876 0 0 0 44460 1264 2997 23 13 22 41 2 2 0 3824668 292936 2791884 0 0 0 25262 1113 4797 28 12 29 31 2 3 0 3784772 292936 2791896 0 0 0 38988 1468 6677 28 12 48 12 2 4 0 3736256 292936 2791904 0 0 0 50970 1530 5217 19 12 49 20 4 2 0 3698056 292936 2791908 0 0 0 43576 1369 7316 20 15 35 30 2 1 0 3667124 292936 2791920 0 0 0 39174 1444 4659 25 16 35 24 6 1 0 3617652 292936 2791928 0 0 0 52430 1347 4681 25 19 20 37 1 3 0 3599992 292936 2790868 0 0 0 40156 1439 4394 20 14 29 37 6 0 0 3797488 292936 2568648 0 0 0 17706 2272 21534 28 23 19 30 0 0 0 3785396 292936 2568736 0 0 0 1156 1237 14057 33 8 0 59 0 0 0 3783568 292936 2568736 0 0 0 704 512 1537 5 2 1 92 1 0 0 3783188 292936 2568752 0 0 0 842 613 1919 6 1 1 92 -anjan -----Original Message----- From: Rod Taylor [mailto:[EMAIL PROTECTED] Sent: Tuesday, October 26, 2004 1:49 PM To: Anjan Dave Cc: Postgresql Performance Subject: RE: [PERFORM] can't handle large number of INSERT/UPDATEs On Tue, 2004-10-26 at 13:42, Anjan Dave wrote: > It probably is locking issue. I got a long list of locks held when we ran select * from pg_locks during a peak time. > > relation | database | transaction | pid | mode | granted > ----------+----------+-------------+-------+------------------+--------- > 17239 | 17142 | | 3856 | AccessShareLock | t How many have granted = false? > Vmstat would show a lot of disk IO at the same time. > > Is this pointing towards a disk IO issue? Not necessarily. Is your IO reaching the limit or is it just heavy? ---------------------------(end of broadcast)--------------------------- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
