On Tue, 12 Aug 2008, Tom Lane wrote:
> This seems like a fairly bad idea, because it introduces a
> gettimeofday() call per transaction.
There's already lots of paths through pgbench that introduce gettimeofday
calls all over the place. I fail to see how this is any different.
> If this were worth doing (which IMHO it isn't)
I think that switching the recommended practice for running pgbench to
something time-based rather than transactions-based would increase the
average quality of results people got considerably. How many times do you
see people posting numbers that worthless because the test ran for a
trivial amount of time? Seems like it happens a lot to me. This patch
was already on my todo list for 8.4 and I'm glad I don't have to write it
myself now.
--
* Greg Smith gsmith@gregsmith.com http://www.gregsmith.com Baltimore, MD