Tom Lane wrote:
> --enable-cassert might have a noticeable performance impact.
> We usually try to not have that on when doing performance testing.
>
All covered in the tool's documentation, and it looks like Kevin did the
right thing during his tests by checking the pg_config output to confirm
the right flags were used. I think we can rule out simple pilot error
here and work under the assumption Kevin found a mild performance
regression under some circumstances with the patch. Hopefully Andres
will be able to replicate the problem, and it sounds like Kevin might be
able to provide more information about it tonight too.
--
Greg Smith 2ndQuadrant Baltimore, MD
PostgreSQL Training, Services and Support
greg@2ndQuadrant.com www.2ndQuadrant.com