Re: Bug #807: Sequence currupted on recovery after kill -9

Florian Weimer <weimer@cert.uni-stuttgart.de>

From: Florian Weimer <Weimer@CERT.Uni-Stuttgart.DE>
To: Darcy Buskermolen <darcy@wavefire.com>
Cc: mthomp@tierfleet.com, pgsql-bugs@postgresql.org
Date: 2002-10-31T20:25:59Z
Lists: pgsql-bugs
Darcy Buskermolen <darcy@wavefire.com> writes:

> I'm sure if you read the footers on Tom's email's it's pretty clearly stated 
> 'do not kill -9 the postmaster'. 
> So I'm not so sure that this is really a bug. 

The operating environment is free to kill -9 the postmaster at any
time, and PostgreSQL is expected to ensure database consistency
nevertheless.

After all, PostgreSQL is a database and not a cardbox.

(But holes in sequence numbers certainly do not endanger database
consistency.)

-- 
Florian Weimer 	                  Weimer@CERT.Uni-Stuttgart.DE
University of Stuttgart           http://CERT.Uni-Stuttgart.DE/people/fw/
RUS-CERT                          fax +49-711-685-5898