On Sun, Mar 29, 2020 at 11:50:10AM -0700, Andres Freund wrote:
Hi,
On March 29, 2020 11:24:32 AM PDT, Alexander Korotkov
<a.korot...@postgrespro.ru> wrote:
clearly a big win on majority
of workloads, I think we still need to investigate different workloads
on different hardware to ensure there is no regression.
Definitely. Which workloads are you thinking of? I can think of those
affected facets: snapshot speed, commit speed with writes, connection
establishment, prepared transaction speed. All in the small and large
connection count cases.
I did measurements on all of those but prepared xacts, fwiw. That
definitely needs to be measured, due to the locking changes around
procarrayaddd/remove.
I don't think regressions besides perhaps 2pc are likely - there's
nothing really getting more expensive but procarray add/remove.
If I get some instructions what tests to do, I can run a bunch of tests
on my machinees (not the largest boxes, but at least something). I don't
have the bandwidth to come up with tests on my own, at the moment.
regards
--
Tomas Vondra http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
