Excessive vacuum times

From: Wes <wespvp(at)syntegra(dot)com>
To: <pgsql-general(at)postgresql(dot)org>
Subject: Excessive vacuum times
Date: 2005-12-12 22:58:51
Message-ID: BFC35CCB.1B2A2%wespvp@syntegra.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Some time ago I reported a problem where right after an index rebuild vacuum
might take a couple of hours, but gets slower and slower at a seeming
increasing rate as time goes on. What takes a couple of ours after a full
index rebuild now takes 12 hours or more.

The problem was determined to be due to the fact that indexes are vacuumed
in index order, not in disk storage order. I don't see anything about this
in the "What's new" for 8.1. Has anything been done to resolve this?

Scheduled index rebuilds are not feasible. I really don't even want to take
a chance on starting one. The database is just a hair under one billion
rows, and could take the entire weekend or more to rebuild.

Wes

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Tony Caduto 2005-12-12 23:04:01 What's new in 8.1.1
Previous Message Tom Lane 2005-12-12 22:42:05 Re: View with an outer join - is there any way to optimise this?