Message-ID: <4ED80128.80600@catalyst.net.nz>
Date: Fri, 02 Dec 2011 11:35:20 +1300
From: Mark Kirkwood <mark.kirkwood@catalyst.net.nz>
User-Agent: Mozilla/5.0 (X11; Linux x86_64;
	rv:8.0) Gecko/20111124 Thunderbird/8.0
MIME-Version: 1.0
To: pgsql-performance@postgresql.org
Subject: Re: Autovacuum Issue
References: 
 <CA+-36iH+UgRP+hhZSOScUtMGTutFtepDNY-E7C_tvubWhbfcjw@mail.gmail.com>
	<CA+TgmobPwvrHXgBYHPcY9jFpaTumGaMHi-vMTm5hFLVxircQGw@mail.gmail.com>
In-Reply-To: 
 <CA+TgmobPwvrHXgBYHPcY9jFpaTumGaMHi-vMTm5hFLVxircQGw@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

On 02/12/11 07:18, Robert Haas wrote:
>
> And also please share your views on my decision about disable autovacuum for
> my application. I am planning to run vacuum command daily on that small
> table which has frequent updates.
> Sounds like a bad plan.
>

If the table has frequent updates vacuuming once a day will not control 
space bloat from dead rows... so your small table's storage will become 
a very large (even though there are only a few undeleted rows), and 
performance will become terrible.

I would suggest tuning autovacuum to wakeup more frequently (c.f 
autovacuum_naptime parameter), so your small table stays small.

Also you didn't mention what version of Postgres you are running. In 8.4 
and later vacuum (hence autovacuum) is much smarter about finding dead 
rows to clean up, and should have less impact. You can also control the 
load autovacuum puts on your system (c.f autovacuum_vacuum_cost_delay 
parameter).

regards

Mark