Re: pg_dumpall failing from possible corrupted shared memory - Mailing list pgsql-general

From Tom Lane
Subject Re: pg_dumpall failing from possible corrupted shared memory
Date
Msg-id 13003.1161871037@sss.pgh.pa.us
Whole thread Raw
In response to Re: pg_dumpall failing from possible corrupted shared memory  (Richard Broersma Jr <rabroersma@yahoo.com>)
Responses Re: pg_dumpall failing from possible corrupted shared memory  (Richard Broersma Jr <rabroersma@yahoo.com>)
List pgsql-general
Richard Broersma Jr <rabroersma@yahoo.com> writes:
> I did not find any coredump files.  Will I need to rebuild postgres to produce coredumps?

No, you just need to make sure the postmaster is started under
"ulimit -c unlimited", not "ulimit -c 0" which is the default
under many Linuxen.  Adding this to the start script is the
easiest way usually.

If you can't get anything but numbers from gdb's stack trace
then you have a "stripped" executable and you will need to
rebuild to get a useful stack trace.

            regards, tom lane

pgsql-general by date:

Previous
From: Devrim GUNDUZ
Date:
Subject: Re: Problem with 8.1.5 RPMs for Redhat AS 4
Next
From: Andrew Sullivan
Date:
Subject: Re: password cookie