On Thu, Apr 25, 2002 at 10:10:16AM +0100, TPCCUVA wrote:
> I'm making a benchmark for to measure the throughput of a database
> system. I'm using postgres 7.1.3 and SQL embedded in C .
>
> The program executes five transacctions with, SELECTs UPDATESs INSERTs,
> DELETEs and FETCHs.
> The workload consists in the intensive execution of this transacctions.
>
> The problem is that the more operations are executed, the more slower
> truns postgres and the transacctions are slower.
Well, when I run lots of queries, sometimes they take a few milliseconds
longer. I think it has to do with my disk wearing out.
Seriously, if you want a detailed explanation you're going to have tell us
exactly what queries you are executing. I have a database up for 3 weeks and
it's just as fast now as when I started it.
What's intensive? 1 per second? Oh, you know about vacuum and vacuum
analyse, right?
--
Martijn van Oosterhout <kleptog@svana.org> http://svana.org/kleptog/
> Canada, Mexico, and Australia form the Axis of Nations That
> Are Actually Quite Nice But Secretly Have Nasty Thoughts About America