Re: Why so slow?

From: Bill Moran <wmoran(at)collaborativefusion(dot)com>
To: "Jim C(dot) Nasby" <jnasby(at)pervasive(dot)com>
Cc: bealach_na_bo(at)hotmail(dot)com, ahodgson(at)simkin(dot)ca, pgsql-performance(at)postgresql(dot)org
Subject: Re: Why so slow?
Date: 2006-05-02 23:28:34
Message-ID: 20060502192834.6c8e9131.wmoran@collaborativefusion.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

"Jim C. Nasby" <jnasby(at)pervasive(dot)com> wrote:

> On Sun, Apr 30, 2006 at 10:03:46AM -0400, Bill Moran wrote:
> > At this point, it seems like you need to do 2 things:
> > 1) Schedule lazy vacuum to run, or configure autovacuum.
> > 2) Schedule some downtime to run "vacuum full" to recover some disk space.
> >
> > #2 only needs done once to get you back on track, assuming that #1 is
> > done properly.
>
> You'll also want to reindex since vacuum full won't clean the indexes
> up. You might also want to read
> http://www.pervasivepostgres.com/instantkb13/article.aspx?id=10087 and
> http://www.pervasivepostgres.com/instantkb13/article.aspx?id=10116.

Reindexing is in a different class than vacuuming. Neglecting to vacuum
creates a problem that gets worse and worse as time goes on. Neglecting
to reindex does not create an infinately growing problem, since empty
index pages are recycled automatically. It's also conceivable that some
usage patterns don't need to reindex at all.

http://www.postgresql.org/docs/8.1/interactive/routine-reindex.html

--
Bill Moran
Potential Technologies
http://www.potentialtech.com

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Tom Lane 2006-05-02 23:30:00 Re: Killing long-running queries
Previous Message Dan Harris 2006-05-02 23:19:52 Killing long-running queries