From: Tom Lane <tgl@sss.pgh.pa.us>
To: Alvaro Herrera <alvherre@alvh.no-ip.org>
cc: Marko Tiikkaja <marko@joh.to>,
 Michael Paquier <michael.paquier@gmail.com>,
        PostgreSQL mailing lists <pgsql-bugs@postgresql.org>
Subject: Re: BUG #14830: Missed NOTIFications, PostgreSQL 9.1.24
In-reply-to: <7833.1507658856@sss.pgh.pa.us>
References: <20171010140621.u4avva4cxylwpyew@alvherre.pgsql>
 <16671.1507645188@sss.pgh.pa.us> <18082.1507647266@sss.pgh.pa.us>
 <7833.1507658856@sss.pgh.pa.us>
Comments: In-reply-to Tom Lane <tgl@sss.pgh.pa.us>
	message dated "Tue, 10 Oct 2017 14:07:36 -0400"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <29601.1507670724.1@sss.pgh.pa.us>
Date: Tue, 10 Oct 2017 17:25:24 -0400
Message-ID: <29602.1507670724@sss.pgh.pa.us>
Precedence: bulk
Sender: pgsql-bugs-owner@postgresql.org

I wrote:
> Concretely, I suggest the attached patch.  I tried Marko's testbed
> against this, and it seems indeed a bit faster than before ---
> running it for 100000 notifies takes about 19.25 seconds, rather than
> 19.5 seconds with HEAD.  But the testbed is probably about the best
> case because it has a bunch of threads sending notifies to one
> receiver, so that the receiver is quite likely to have multiple
> messages to read at a time.  OTOH, in a situation where the NOTIFY
> traffic isn't so high, it probably hardly matters anyway.

It occurred to me that it's easy to measure the worst case, ie always
one message per notify interrupt:

$ cat bnch.txt
LISTEN foo\; NOTIFY foo;
$ pgbench -n -c 1 -T 10 -f bnch.txt

On this case it seems that the patch is circa 2% slower than HEAD,
though that's only barely above the noise level in my measurements.
I think that's probably acceptable.

			regards, tom lane


-- 
Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs