On Fri, Oct 01, 2004 at 10:10:40AM -0700, Josh Berkus wrote:
> Transparent "query caching" is the "industry standard" for how these things
> are handled. However, Postgres' lack of this feature has made me consider
> other approaches, and I'm starting to wonder if the "standard" query caching
> -- where a materialized query result, or some reduction thereof, is cached in
> database memory -- isn't the best way to cache things. I'm going to
> abbreviate it "SQC" for the rest of this e-mail.
Not to quibble, but are you sure that's the standard? Oracle and DB2
don't do this, and I didn't think MSSQL did either. What they do do is
cache query *plans*. This is a *huge* deal in Oracle; search
http://asktom.oracle.com for 'soft parse'.
In any case, I think a means of marking some specific queries as being
cachable is an excellent idea; perfect for 'static data' scenarios. What
I don't know is how much will be saved.
--
Jim C. Nasby, Database Consultant decibel@decibel.org
Give your computer some brain candy! www.distributed.net Team #1828
Windows: "Where do you want to go today?"
Linux: "Where do you want to go tomorrow?"
FreeBSD: "Are you guys coming, or what?"