From: | Alvaro Herrera <alvherre(at)commandprompt(dot)com> |
---|---|
To: | Harald Fuchs <hari(dot)fuchs(at)gmail(dot)com> |
Cc: | pgsql-admin(at)postgresql(dot)org |
Subject: | Re: autovacuum question |
Date: | 2009-02-05 18:27:37 |
Message-ID: | 20090205182737.GD2978@alvh.no-ip.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Harald Fuchs escribió:
> In article <20090204165933(dot)GB32680(at)alvh(dot)no-ip(dot)org>,
> Alvaro Herrera <alvherre(at)commandprompt(dot)com> writes:
>
> > Marc Mamin wrote:
> >> There are 2 features that would be helpfull in my case:
> >>
> >> - define vacuum properties on table groups rather than on given tables:
> > Hmm. I'm not really sure how could we handle this kind of situation.
>
> Couldn't we use schemas for that? Put all those tables into a
> separate schema and somehow tell autovacuum to leave that schema
> alone.
I don't like that. It would force the users to use schemas in a certain
way, taking flexibility away from them.
In an old design for autovacuum "maintenance windows" (which were
promised for 8.2 and still haven't seen the light of day) we were
introducing "table groups" as a concept. Sounds like something that
applies here too. If you can attach a table to a group during creation,
you can decide its autovacuum schedule right there. You should be able
to do something like
create table d_123 ( ... ) WITH (autovacuum_group = 'dont_ever_vacuum');
which means autovacuum automatically ignores that table.
--
Alvaro Herrera http://www.CommandPrompt.com/
PostgreSQL Replication, Consulting, Custom Development, 24x7 support
From | Date | Subject | |
---|---|---|---|
Next Message | Simon Riggs | 2009-02-05 20:53:35 | Re: Totally inappropriate by-line for a major representative of this organization |
Previous Message | Sylvain Avril | 2009-02-05 16:57:43 | Re: Totally inappropriate by-line for a major representative of this organization |