Re: vacuumdb question/problem - Mailing list pgsql-admin

From Kevin Grittner
Subject Re: vacuumdb question/problem
Date
Msg-id 4E204C0B020000250003F3AB@gw.wicourts.gov
Whole thread Raw
In response to Re: vacuumdb question/problem  (David Ondrejik <David.Ondrejik@noaa.gov>)
List pgsql-admin
David Ondrejik <David.Ondrejik@noaa.gov> wrote:

> So I had to kill my process, recover disk space and get the
> machine back in working condition for the weekend. I guess I will
> attempt to do the full vacuum again next week.

Why do you think that you need a VACUUM FULL?  That is only needed
as an extreme measure in rather unusual circumstances.  We run
hundreds of PostgreSQL databases, and have not used VACUUM FULL for
years.  On the other hand, we VACUUM ANALYZE most databases in their
entirety every night.  (Be very careful not to confuse VACUUM FULL
with vacuum of an entire database.)

On the other hand, the failed attempt at VACUUM FULL may have
bloated your indexes, so you might want to REINDEX the table next
week, followed by a VACUUM ANALYZE (without the FULL).

-Kevin

pgsql-admin by date:

Previous
From: Tom Lane
Date:
Subject: Re: vacuumdb question/problem
Next
From: Isabella Ghiurea
Date:
Subject: PG 8.4 and identity columns