Kevin Brown <kevin@sysexperts.com> writes:
> Tom Lane wrote:
>> ... But those same
>> limitations cause the results to be unstable from run to run, which
>> is why I don't have a lot of faith in reports of pgbench numbers.
>> You need to work quite hard to get reproducible numbers out of it.
> The interesting question is whether that's more an indictment of how
> PG does things or how pg_bench does things.
I didn't draw a conclusion on that ;-). I merely pointed out that the
numbers are unstable, and therefore not to be trusted without a great
deal of context ...
regards, tom lane