Re: Postgres 8.4 memory related parameters

Shaun Thomas <sthomas(at)peak6(dot)com> wrote:

> So with a dual X5675, that's 12 cores. My numbers peaked at
> 24-concurrency. At that concurrency, HT was 60% faster than
> non-HT. Sorry if I mixed my terminology. :)

No problem -- I appreciate the information. I just wanted to be
sure I was understanding it correctly. So, with hyperthreading
turned on, the optimal number of active connections was twice the
actual cores. And since the active data set was fully cached, disk
spindles were not a resource which played any significant role in
the test, making the "effective spindle count" zero. So this is one
more data point confirming the overall accuracy of the formula I
use, and providing evidence that it is not affected by use of
hyperthreading if you base your numbers on actual cores.

optimal pool size = ((2 * actual core count) + effective spindle
count)

optimal pool size = ((2 * 12) + 0)

optimal pool size = 24

Thanks!

-Kevin