Re: Database corruption in 7.0.3 - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Database corruption in 7.0.3
Date
Msg-id 13737.984669903@sss.pgh.pa.us
Whole thread Raw
In response to Database corruption in 7.0.3  (Tim Allen <tim@proximity.com.au>)
List pgsql-hackers
Tim Allen <tim@proximity.com.au> writes:
> Are there any known database-corrupting bugs in 7.0.3?

None that aren't also in earlier releases, AFAIR, so your report is
fairly troubling.  However there's not enough here to venture a guess
about the source of the problem.

Do you see any backend crashes or other misbehavior before the VACUUM
error pops up, or is that the only symptom?

It would be a good idea to rebuild the system with assert checks on
(configure --enable-cassert), in hopes that some Assert a little closer
to the source of the problem will fire.  Also, if you can spare some
disk space for logging, running the postmaster with -d2 to log all
queries might provide useful historical context when the problem
reappears.

I would like to be able to study the corrupted table, as well.  Can you
see your way to either giving me access to your machine, or (if the
database isn't too large) sending me a tar dump of the whole $PGDATA
directory next time it happens?

Please contact me off-list so we can figure out how best to pursue this
problem.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: PostgreSQL on multi-CPU systems
Next
From: Tom Lane
Date:
Subject: Re: Sheduling in SQL