From: | "Dave Golombek" <daveg(at)blackducksoftware(dot)com> |
---|---|
To: | "'Tom Lane'" <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: Problem with inherited tables vs query planning |
Date: | 2007-05-18 14:34:48 |
Message-ID: | 20070518143534.5CE615C4512@smtp.blackducksoftware.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Tom Lane writes:
> "Dave Golombek" <daveg(at)blackducksoftware(dot)com> writes:
> > Is there a way I can reformulate the query to help the planner use the
> > indices?
>
> Use 8.2. Also put an index on the base table, not only the children ---
> the forced seqscan on the base weighs down the cost estimate for the
> plan you would like to have. (With sufficiently large child tables,
> that might not matter, but it sure does for this toy example.)
Ah, I forgot to try the index on the base table using 8.2, which does indeed
solve the problem. It unfortunately doesn't help with 8.1.4, which we have
in the field; any thoughts on workarounds for older versions or should I
just use a function until we can upgrade everywhere? We have 150 million
rows spread across 16 child tables, which should help cost estimation.
Thanks,
Dave
From | Date | Subject | |
---|---|---|---|
Next Message | Andrew Sullivan | 2007-05-18 14:45:22 | Re: Data replication through disk replication |
Previous Message | Andrew Sullivan | 2007-05-18 14:29:03 | Re: Fault Tolerant Postgresql (two machines, two postmasters, one disk array) |