Re: Shortening time of vacuum analyze

From: Francisco Reyes <lists(at)natserv(dot)com>
To: Andrew Sullivan <andrew(at)libertyrms(dot)info>
Cc: PostgreSQL general list <pgsql-general(at)postgresql(dot)org>
Subject: Re: Shortening time of vacuum analyze
Date: 2002-01-30 22:12:53
Views: Raw Message | Whole Thread | Download mbox | Resend email
Lists: pgsql-general

On Wed, 30 Jan 2002, Andrew Sullivan wrote:

> On Wed, Jan 30, 2002 at 12:23:50PM -0500, Francisco Reyes wrote:
> > You can put the delete/load inside a transaction so the users will never
> > see an empty table. Truncate can not be placed inside a transaction.
> True enough. But why not leave the table unvacuumed, then, until a
> more convenient time? You are, of course, paying a cost in
> performance during that time, but not as great as you would with
> vacuum.

We do the vacuum right after the load of all the new tables, but before we
do all our daily reports. An operation that takes 30 minutes if all is
vacuumed takes about 2 hours if not vacuumed.

Given that 7.2 is due out shortly I have comed up with some other ways of
improving time.. ie delaying re-freshing some data which rarely ever
changes until 7.2 is out.

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Marc Munro 2002-01-31 00:55:06 Row-based security (virtual private database)
Previous Message Tom Lane 2002-01-30 21:20:21 Re: process exited with status 11 after XLogFlush: request is not satisfied