From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | "Christopher Kings-Lynne" <chriskl(at)familyhealth(dot)com(dot)au> |
Cc: | "Brett Toolin" <brett(at)familyhealth(dot)com(dot)au>, "Hackers" <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Postgres 7.2.2 Segment Error |
Date: | 2002-09-20 14:24:19 |
Message-ID: | 16030.1032531859@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
"Christopher Kings-Lynne" <chriskl(at)familyhealth(dot)com(dot)au> writes:
> The sessions table holds normal site session data, like a uid, username,
> some other stuff, etc. However entries older than two hours or so get
> deleted. We VACUUM everynight, so why is the on-disk relation growing so
> huge?
FSM not big enough, perhaps? Try doing a vacuum full, then looking to
see how big the table is (in physical blocks) after one day's normal
usage. You need at least enough FSM space for that many blocks
... unless you want to vacuum it more often.
> However, I cannot repeat the error now.
If you can't reproduce the error then I'm pretty well convinced that
there is no problem in the stored data itself. This was either a
hardware glitch or a software bug causing a memory stomp on the top byte
of an item pointer retrieved from the index. Although I can't rule out
the latter, I find it unlikely given that we don't have similar reports
from other people.
You may as well do the VACUUM FULL --- I doubt we can learn anything
from examining the table.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Dave Page | 2002-09-20 14:27:47 | Re: Where to post a new PostgreSQL utility? |
Previous Message | Lee Kindness | 2002-09-20 14:18:49 | Re: Where to post a new PostgreSQL utility? |