thread-index: AccS3YRfPKg45pnVQ8WY2LAlJCbp+Q==
Message-ID: <456C189F.7080103@enterprisedb.com>
Date: Tue, 28 Nov 2006 11:08:15 +0000
Content-Class: urn:content-classes:message
From: "Heikki Linnakangas" <heikki@enterprisedb.com>
Importance: normal
User-Agent: Thunderbird 1.5.0.7 (X11/20060927)
MIME-Version: 1.0
To: "Michael Simms" <michael@tuxgames.com>
Cc: <pgsql-bugs@postgresql.org>
Subject: Re: BUG #2784: Performance serious degrades over a period of a month
References: <200611261635.kAQGZqKE025405@wwwmaster.postgresql.org>
In-Reply-To: <200611261635.kAQGZqKE025405@wwwmaster.postgresql.org>
Content-Type: text/plain;
	format=flowed;
	charset="ISO-8859-1"
Content-Transfer-Encoding: 7bit

Michael Simms wrote:
> The following bug has been logged online:
> 
> Bug reference:      2784
> Logged by:          Michael Simms
> Email address:      michael@tuxgames.com
> PostgreSQL version: 8.1.4
> Operating system:   Linux kernel 2.6.12
> Description:        Performance serious degrades over a period of a month
> Details: 
> 
> OK, we have a database that runs perfectly well after a dump and restore,
> but over a period of a month or two, it just degrades to the point of
> uselessness.
> vacuumdb -a is run every 24 hours. We have also run for months at a time
> using -a -z but the effect doesnt change.

You might have a hung transaction that never finishes open, which 
prevents vacuum from removing old tuple versions. Or you might have too 
low FSM settings as others suggested.

I'd try running VACUUM VERBOSE by hand, and taking a good look at the 
output. If there's nothing obviously wrong with it, please send the 
output back to the list (or pgsql-performance, as Tom suggested), and 
maybe we can help.

-- 
   Heikki Linnakangas
   EnterpriseDB   http://www.enterprisedb.com