Re: 8.3devel slower than 8.2 under read-only load

On Wed, 2007-11-21 at 20:04 +0100, Guillaume Smet wrote:

> number of clients: 1
> number of transactions per client: 1000
> number of transactions actually processed: 1000/1000
> tps = 784.819087 (including connections establishing)
> tps = 786.884214 (excluding connections establishing)
>
> All the queries are index scans (SELECT a couple of fields FROM a
> table WHERE one or two conditions). I checked the plans on both 8.2
> and 8.3 and they are identical. I made several runs and numbers are
> consistent.

Please do tests of at least 2 minutes duration. A 1.25 second test isn't
enough. Please confirm you have VACUUM ANALYZED each db beforehand.

Have you checked that the EXPLAIN ANALYZEs are essentially identical
also? Is the data identical on both systems?

How do the postgresql.conf files differ?

Please find out any differences you can, so we can rule things out.

Also, do a run with SELECT 1 FROM table where col = constant; so we can
assess differences in path without cache or data being relevant.

--
Simon Riggs
2ndQuadrant http://www.2ndQuadrant.com