| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Thomas Munro <thomas(dot)munro(at)gmail(dot)com> |
| Cc: | Nathan Bossart <nathandbossart(at)gmail(dot)com>, Kyotaro Horiguchi <horikyota(dot)ntt(at)gmail(dot)com>, alvherre(at)alvh(dot)no-ip(dot)org, bharath(dot)rupireddyforpostgres(at)gmail(dot)com, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: Suppressing useless wakeups in walreceiver |
| Date: | 2022-11-13 22:08:04 |
| Message-ID: | 742545.1668377284@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Thomas Munro <thomas(dot)munro(at)gmail(dot)com> writes:
> And with that change and a pgindent, pushed.
There is something very seriously wrong with this patch.
On my machine, running "make -j10 check-world" (with compilation
already done) has been taking right about 2 minutes for some time.
Since this patch, it's taking around 2:45 --- I did a bisect run
to confirm that this patch is where it changed.
The buildfarm is showing a hit, too. Comparing the step runtimes at
https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=longfin&dt=2022-11-08%2005%3A29%3A28
https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=longfin&dt=2022-11-08%2007%3A49%3A31
it'd seem that most tests involving walreceivers got much slower:
pg_basebackup-check from 00:29 to 00:39,
pg_rewind-check went from 00:56 to 01:26,
and recovery-check went from 03:56 to 04:45.
Curiously, subscription-check only went from 03:26 to 03:29.
I've not dug into it further than that, but my bet is that some
required wakeup condition was not accounted for.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Thomas Munro | 2022-11-13 22:18:13 | Re: Suppressing useless wakeups in walreceiver |
| Previous Message | Thomas Munro | 2022-11-13 21:28:14 | Re: Reducing power consumption on idle servers |