| From: | Gregory Stark <stark(at)enterprisedb(dot)com> |
|---|---|
| To: | "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
| Cc: | "Gregory Stark" <gsstark(at)mit(dot)edu>, "Heikki Linnakangas" <heikki(at)enterprisedb(dot)com>, <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Function execution costs 'n all that |
| Date: | 2007-01-16 16:46:53 |
| Message-ID: | 87wt3mly0y.fsf@stark.xeocode.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
"Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us> writes:
> Gregory Stark <gsstark(at)mit(dot)edu> writes:
>> I imagine you've thought of this already but just in case, the cost of the
>> function call has to be combined with the selectivity to get this right. If
>> you can do an expensive but very selective clause first and save 100 cheap
>> calls that almost always return true it may still be worthwhile.
>
> I've thought of it, but I haven't figured out a reasonable algorithm for
> ordering the clauses in view of that. Have you?
Hum, I hadn't tried. Now that I think about it it's certainly not obvious.
And picturing the possible failure modes I would rather it execute cheap
expressions more often than necessary than call some user-defined perl
function that could be doing i/o or involve waiting on other resources any
more than absolutely necessary. So I guess what you originally described is
safest.
--
Gregory Stark
EnterpriseDB http://www.enterprisedb.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Enrico | 2007-01-16 17:04:07 | Index for similarity search |
| Previous Message | Andrew Dunstan | 2007-01-16 16:20:51 | Re: [HACKERS] Checkpoint request failed on version 8.2.1. |