mlw <markw(at)mohawksoft(dot)com> writes:
> I have been doing some benchmarking on RedHat 7.2
I think you neglected to mention the pgbench scale factor?
> Journallng always affects performance. This is no surprise. If you
> have fsync enabled, the affect is less pronounced. (this is also no
> surprise). One interesting thing, as the number of concurrent
> connections goes up, the impact of journalling and fsync are reduced.
If you are approaching the scale factor then that just means that the
backends are spending too much CPU on contending for row locks ...
> I'm not sure of the digestion of all these numbers, but I thought some of you
> guys would be interested in comparing notes.
I find it easier to digest graphs than numbers, so here are a couple of
GIFs of Mark's results. The first is the filesystem/fsync comparison,
the second the NBuffers comparison.
regards, tom lane
