On 2016-05-03 11:12:03 -0500, Kevin Grittner wrote:
> On Tue, May 3, 2016 at 10:58 AM, Robert Haas <robertmhaas@gmail.com> wrote:
>
> > - Snapshot Too Old. Tom, Andres, and Bruce want this reverted.
> > It regresses performance significantly when turned on.
>
> When turned on, it improves performance in some cases and regresses
> performance in others. Don't forget it is currently back-patched
> to 9.4 and in use for production by users who could not use
> PostgreSQL without the feature. PostgreSQL failed their
> performance tests miserably without the feature, and passes with
> it.
>
> > It originally regressed performance significantly even when
> > turned off,
>
> Which was wildly exaggerated since most of the benchmarks
> purporting to show that actually had it turned on. I don't think
> the FUD from that has really evaporated.
Oh, ffs. The first report of the whole issue was *with default
parameters*:
http://archives.postgresql.org/message-id/CAPpHfdtMONZFOXSsw1HkrD9Eb4ozF8Q8oCqH4tkpH_girJPPuA%40mail.gmail.com
The issue with 0 v. -1 (and 0 vs. > 0 makes a big performance
difference, so it's not that surprising to interpret numbers that way)
was immediately addressed by another round of benchmarks after you
pointed it out.
> > but that might be fixed now.
>
> Certainly all evidence suggests that, FUD to the contrary.
So it's now FUD to report issues with a patch that obviously hasn't
received sufficient benchmarking? Give me break.
Andres