Re: Serious Crash last Friday

On Wed, Jun 19, 2002 at 11:35:19AM +0200, Henrik Steffen wrote:
> Hello all,
>
> unfortunately I didn't get a really helping answer from the list yet.

We are having trouble helping, because you are not providing the
information that people have asked you for.

> Additionally yesterday night there was again a problem with some SELECTs:
>
> NOTICE:  Message from PostgreSQL backend:
> The Postmaster has informed me that some other backend

As someone else already told you, this is no help because it's a
message about some _other_ back end.  You need to turn on debugging.
Find out where the problem is:

-Track the PIDs
-When this happens, find the PID that is causing the corruption
-Then you can trace further back in your log file to find the query
that was running when it happened.

If the backend is dying abnormally, it should also leave a core file
around, on which you can use a debugger.

> This is really annoying.

But no-one can help you if you don't follow the proffered advice for
how to help.

> When I noticed it this morning, I dropped all indexes and recreated them.
> Then I ran a VACUUM FULL VERBOSE ANALYZE - afterwards the same query worked
> properly again.

The query in question is _not_ the problem.  The error message tells
you so: some _other_ back end is dying.

A

--
----
Andrew Sullivan                               87 Mowat Avenue
Liberty RMS                           Toronto, Ontario Canada
<andrew@libertyrms.info>                              M6K 3E3
                                         +1 416 646 3304 x110