Date: Tue, 31 Jan 2023 16:05:21 -0800
From: Nathan Bossart <nathandbossart@gmail.com>
To: Amit Kapila <amit.kapila16@gmail.com>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Thomas Munro <thomas.munro@gmail.com>,
	"Hayato Kuroda (Fujitsu)" <kuroda.hayato@fujitsu.com>,
	"pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Subject: Re: suppressing useless wakeups in logical/worker.c
Message-ID: <20230201000521.GA3147405@nathanxps13>
References: 
 <CA+hUKGLpH=y4=Zx+=oLE8WvLqLHWNC-PEDARrrt=x=J_REL9Fg@mail.gmail.com>
 <2888102.1674705840@sss.pgh.pa.us>
 <3126727.1674759248@sss.pgh.pa.us>
 <20230126194812.GA1702315@nathanxps13>
 <3165443.1674763470@sss.pgh.pa.us>
 <20230126202301.GA1703463@nathanxps13>
 <3211341.1674767391@sss.pgh.pa.us>
 <20230126212255.GA1746118@nathanxps13>
 <3220831.1674772625@sss.pgh.pa.us>
 <CAA4eK1K6nPBk_fXGUSiUt=vJE__+Xxy+fc4jjciWgjK7b2441A@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <CAA4eK1K6nPBk_fXGUSiUt=vJE__+Xxy+fc4jjciWgjK7b2441A@mail.gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/20230201000521.GA3147405%40nathanxps13>
Precedence: bulk

On Sat, Jan 28, 2023 at 10:26:25AM +0530, Amit Kapila wrote:
> On Fri, Jan 27, 2023 at 4:07 AM Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> Returning to the prior patch ... I don't much care for this:
>>
>> +                    /* Maybe there will be a free slot in a second... */
>> +                    retry_time = TimestampTzPlusSeconds(now, 1);
>> +                    LogRepWorkerUpdateSyncStartWakeup(retry_time);
>>
>> We're not moving the goalposts very far on unnecessary wakeups if
>> we have to do that.  Do we need to get a wakeup on sync slot free?
>> Although having to send that to every worker seems ugly.  Maybe this
>> is being done in the wrong place and we need to find a way to get
>> the launcher to handle it.

It might be feasible to set up a before_shmem_exit() callback that wakes up
the apply worker (like is already done for the launcher).  I think the
apply worker is ordinarily notified via the tablesync worker's notify_pid,
but AFAICT there's no guarantee that the apply worker hasn't restarted with
a different PID.

> + /*
> + * Since process_syncing_tables() is called conditionally, the
> + * tablesync worker start wakeup time might be in the past, and we
> + * can't know for sure when it will be updated again.  Rather than
> + * spinning in a tight loop in this case, bump this wakeup time by
> + * a second.
> + */
> + now = GetCurrentTimestamp();
> + if (wakeup[LRW_WAKEUP_SYNC_START] < now)
> + wakeup[LRW_WAKEUP_SYNC_START] =
> TimestampTzPlusSeconds(wakeup[LRW_WAKEUP_SYNC_START], 1);
> 
> Do we see unnecessary wakeups without this, or delay in sync?

I haven't looked too cloesly to see whether busy loops are likely in
practice.

> BTW, do we need to do something about wakeups in
> wait_for_relation_state_change()?

... and wait_for_worker_state_change(), and copy_read_data().  From a quick
glance, it looks like fixing these would be a more invasive change.  TBH
I'm beginning to wonder whether all this is really worth it to prevent
waking up once per second.

-- 
Nathan Bossart
Amazon Web Services: https://aws.amazon.com