From: | Greg Stark <gsstark(at)mit(dot)edu> |
---|---|
To: | PFC <lists(at)peufeu(dot)com> |
Cc: | "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-performance(at)postgresql(dot)org, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: [HACKERS] Big IN() clauses etc : feature proposal |
Date: | 2006-05-09 15:00:29 |
Message-ID: | 8764kfz1iq.fsf@stark.xeocode.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers pgsql-performance |
PFC <lists(at)peufeu(dot)com> writes:
>
> I really like this. It's clean, efficient, and easy to use.
>
> This would be a lot faster than using temp tables.
> Creating cursors is very fast so we can create two, and avoid doing
> twice the same work (ie. hashing the ids from the results to grab categories
> only once).
Creating cursors for a simple plan like a single sequential scan is fast
because it's using the original data from the table. But your example was
predicated on this part of the job being a complex query. If it's a complex
query involving joins and groupings, etc, then it will have to be materialized
and there's no (good) reason for that to be any faster than a temporary table
which is effectively the same thing.
--
greg
From | Date | Subject | |
---|---|---|---|
Next Message | Mitchell Skinner | 2006-05-09 15:58:05 | Re: [HACKERS] Big IN() clauses etc : feature proposal |
Previous Message | Tom Lane | 2006-05-09 13:56:01 | Re: current version: Patch - Have psql show current values |
From | Date | Subject | |
---|---|---|---|
Next Message | Bruno Wolff III | 2006-05-09 15:18:48 | Re: Arguments Pro/Contra Software Raid |
Previous Message | Steve Atkins | 2006-05-09 14:41:16 | Re: Arguments Pro/Contra Software Raid |