| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | "Christopher Kings-Lynne" <chriskl(at)familyhealth(dot)com(dot)au> |
| Cc: | "Hackers" <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Postgres 7.2.2 Segment Error |
| Date: | 2002-09-19 14:12:59 |
| Message-ID: | 25581.1032444779@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
"Christopher Kings-Lynne" <chriskl(at)familyhealth(dot)com(dot)au> writes:
> I just saw this in my logs:
> 2002-09-18 12:13:10 ERROR: cannot open segment 1 of relation users_sessions
> (target block 1342198864): No such file or directory
> This query caused it:
> DELETE FROM users_sessions WHERE changed < ('now'::timestamp - '1440
> minutes'::interval) AND name = 'fhnid';
What does EXPLAIN show as the plan for that query? I'm guessing an
indexscan, and that the error was caused by reading a broken item
pointer from the index. (1342198864 = hex 50005450, which sure looks
like the upper 5 shouldn't be there ... how big is the table, anyway?)
> However, I cannot repeat the error now. Is this a bug in postgres
> somewhere.
If the broken item pointer were indeed in the index, I'd expect it to be
100% repeatable. I'm wondering about flaky memory or some such. Have
you run any hardware diagnostics?
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Zeugswetter Andreas SB SD | 2002-09-19 14:57:30 | Re: Proposal for resolving casting issues |
| Previous Message | Tom Lane | 2002-09-19 13:53:58 | Re: Proposal for resolving casting issues |