Re: pgbench progress report improvements - split 3 v2 - A

On Sun, Oct 06, 2013 at 05:04:56PM +0200, Fabien COELHO wrote:
> > Patch (2): Make --initialize mode respect --progress.
> >Rejected
>
> I missed this one...

See the second half of this message, including quoted material:
http://www.postgresql.org/message-id/CA+TgmoZNXkm-EtszHX=KWq34H5Ni4CS8DG31no86cmDryAqZ_w@mail.gmail.com

> > Patch (5): Take thread start time at the beginning of the thread.
> >Returned with Feedback
>
> Hmmm. I fought back on the feedback:-) I thought my arguments where
> good enough to consider an accept.

Here is the feedback in question:
http://www.postgresql.org/message-id/20130930223621.GA125986@tornado.leadboat.com

With or without the patch, reported performance figures are uninformative when
thread start time is substantial relative to benchmark duration. A mere time
accounting change will not help much; improving this requires tighter
synchronization around the start of actual query activity across threads. I
didn't read anything in your response as disagreement with that conclusion.

nm

--
Noah Misch
EnterpriseDB http://www.enterprisedb.com