Re: performance issue with a specific query

Hi!

thanks for the quick help
i managed to reduce the response time from seemingly taking infinity to a
tolerable level, which is by the way a huge improvement.
What helped was the functional index on date_trunc('day',yourfieldhere) as
Scott suggested.
I tried to disable the geqo, but it didn't make any noticeable difference.
For now, I am happy with the result, if the result set stays the same I can
live without an upgrade.

> 1: Which 8.xx? 8.0.x or 8.1.x? 8.1.x is literally light years ahead
> of 7.4 in terms of performance. 8.0 is somewhere between them. The
> performance difference you're seeing is pretty common.

The benchmark was done on a small notebook running 8.1.4.1, versus 7.4.13 on
a 2gig P4 server. The difference is astounding, even without the functional
index 8.1 is 5-10 times faster than a full fledged server.

3: You are NOT Stuck on 7.4.13. I have a RHEL server that will be
> running 8.1.4 or so pretty soon as a dataware house. It may get updated
> to RHEL4, may not. You can either compile from the .tar.[gz|bz2] files
> or download the PGDG rpms for your distro.

I know, but that's was I was trying to avoid. It is not that I would use the
RHEL support provided for 7.4.13, but you know, staying official is the
whole point of subscribing to RHEL4.

Moreover, since many of our other applications are running happily running
under 7.4, I would be afraid to upgrade the whole thing.

So, again, thanks everybody for the help, you saved the day for me.

Regards
Eliott