From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Sean Shanny <shannyconsulting(at)earthlink(dot)net> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: An out of memory error when doing a vacuum full |
Date: | 2003-12-29 18:50:41 |
Message-ID: | 16093.1072723841@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Sean Shanny <shannyconsulting(at)earthlink(dot)net> writes:
> sort_mem = 64000 # min 64, size in KB
You might want to lower that; a complex query could easily use several
times sort_mem. Whether this is the immediate source of your problem
with the other query is hard to tell.
> vacuum_mem = 32767 # min 1024, size in KB
That seems all right, but I recollect now that it only applies to plain
VACUUM not VACUUM FULL. VACUUM FULL needs to keep track of *all* the
free space in a table, and so it's certainly possible that vacuuming a
huge table with many dead tuples could require lots of memory. I can't
recall anyone else ever complaining about VACUUM FULL running out of
memory, though, so there may be some other contributing factor in your
situation. Too bad you reloaded the table --- it would be interesting
to see if increasing your 512Mb datasize ulimit would have allowed the
VACUUM FULL to complete. (Not but what it would've taken forever :-()
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2003-12-29 19:04:04 | Re: Out of memory error when doing an update with IN clause |
Previous Message | Andreas | 2003-12-29 18:06:16 | Re: simple auto-updating timestamp ? |