| From: | Thomas Munro <thomas(dot)munro(at)gmail(dot)com> |
|---|---|
| To: | Nicola Contu <nicola(dot)contu(at)gmail(dot)com> |
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, "Ray O'Donnell" <ray(at)rodonnell(dot)ie>, "pgsql-generallists(dot)postgresql(dot)org" <pgsql-general(at)lists(dot)postgresql(dot)org>, Alessandro Aste <Alessandro(dot)aste(at)gtt(dot)net> |
| Subject: | Re: Performance comparison between Pgsql 10.5 and Pgsql 11.2 |
| Date: | 2019-03-04 20:54:23 |
| Message-ID: | CA+hUKGJXQ8Zg3Vy66-U7iCDjRqjkcH11+yCXxa=MEwLq+HZoAg@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
On Tue, Mar 5, 2019 at 5:01 AM Nicola Contu <nicola(dot)contu(at)gmail(dot)com> wrote:
> Attached a part of the strace running the pgbench command for pg11
> Also attached strace_10 for pg10.6.
That looks like strace output from pgbench, and I don't see any
interesting differences between v10 and v11 (though I'm surprised to
see it using poll() instead of ppoll(), and opening a new connection
for every transaction).
How about syscalls on the server side? You could start it with
something like "strace -f path/to/postgres -D path/to/pgdata" (-f for
follow children), and perhaps also use -c so that it shows aggregated
data (up until you ^C it) instead of every syscall?
--
Thomas Munro
https://enterprisedb.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Perumal Raj | 2019-03-04 21:37:30 | Re: Question about pg_upgrade from 9.2 to X.X |
| Previous Message | Jeff Janes | 2019-03-04 20:03:33 | Re: Optimizing Database High CPU |