From: | Tarhon-Onu Victor <mituc(at)iasi(dot)rdsnet(dot)ro> |
---|---|
To: | Dror Matalon <dror(at)zapatec(dot)com> |
Cc: | Postgresql Performance <pgsql-performance(at)postgresql(dot)org> |
Subject: | Re: Various performance questions |
Date: | 2003-10-27 15:15:05 |
Message-ID: | Pine.LNX.4.58.0310271708470.8898@blackblue.iasi.rdsnet.ro |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
On Sun, 26 Oct 2003, Dror Matalon wrote:
> Here's the structure of the items table
[snip]
> pubdate | timestamp with time zone |
> Indexes:
> "item_channel_link" btree (channel, link)
> "item_created" btree (dtstamp)
> "item_signature" btree (signature)
> "items_channel_article" btree (channel, articlenumber)
> "items_channel_tstamp" btree (channel, dtstamp)
>
>
> 5. Any other comments/suggestions on the above setup.
Try set enable_seqscan = off; set enable_indexscan = on; to
force the planner to use one of the indexes. Analyze the queries from
your application and see what are the most used columns in WHERE clauses
and recreate the indexes. select count(*) from items where channel <
5000; will never use any of the current indexes because none matches
your WHERE clause (channel appears now only in multicolumn indexes).
--
Any views or opinions presented within this e-mail are solely those of
the author and do not necessarily represent those of any company, unless
otherwise expressly stated.
From | Date | Subject | |
---|---|---|---|
Next Message | Christopher Kings-Lynne | 2003-10-27 15:16:05 | Linux Filesystem Shootout |
Previous Message | Greg Stark | 2003-10-27 15:09:09 | Re: Various performance questions |