| From: | Neil Conway <neilc(at)samurai(dot)com> |
|---|---|
| To: | "scott(dot)marlowe" <scott(dot)marlowe(at)ihs(dot)com> |
| Cc: | Rajesh Kumar Mallah <mallah(at)trade-india(dot)com>, David Shadovitz <david(at)www(dot)shadovitz(dot)com>, "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org> |
| Subject: | Re: COUNT & Pagination |
| Date: | 2004-01-19 23:58:44 |
| Message-ID: | 8765f7ij7v.fsf@mailbox.samurai.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
"scott.marlowe" <scott(dot)marlowe(at)ihs(dot)com> writes:
> Yes, previously run query should be faster, if it fits in kernel
> cache.
Or the PostgreSQL buffer cache.
> Plus, the design of Postgresql is such that it would have to do a
> LOT of cache checking to see if there were any updates to the
> underlying data between selects.
Last I checked (which was a while ago, admittedly), the MySQL design
completely purges the query cache for a relation whenever that
relation is mentioned in an INSERT, UPDATE, or DELETE. When this was
discussed (check the -hackers archives for more), IIRC the consensus
was that it's not worth implementing it if we can't do better than
that.
-Neil
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Neil Conway | 2004-01-20 00:01:01 | Re: Trigger question |
| Previous Message | Tom Lane | 2004-01-19 02:21:00 | Re: Join optimisation Quandry |