ietf-nntp Re: INN 2.x-CURRENT against a Cyclone server...

[Stan O. Barber]

rra at stanford.edu is not on the ietf-nntp-post list, so I am forwarding this.

-----Original Message-----


>From sob at announcer.academ.com  Mon Jan 25 11:55:20 1999
Received: from academ.com (root at ACADEM.COM [198.137.249.2])
	by announcer.academ.com (8.9.1/8.9.0) with ESMTP id LAA27347
	for <ietf-nntp at ANNOUNCER.ACADEM.COM>; Mon, 25 Jan 1999 11:55:20 -0600 (CST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.12.23])
	by academ.com (8.9.1/8.9.0) with SMTP id LAA01725
	for <ietf-nntp at academ.com>; Mon, 25 Jan 1999 11:55:19 -0600 (CST)
Received: (qmail 5905 invoked by uid 50); 25 Jan 1999 17:55:07 -0000
To: inn-workers at isc.org, ietf-nntp at academ.com
Subject: Re: INN 2.x-CURRENT against a Cyclone server...
References: <199901251609.IAA28881 at karoshi.ucsd.edu>
From: Russ Allbery <rra at stanford.edu>
In-Reply-To: Brian Kantor's message of "Mon, 25 Jan 1999 08:09:18 -0800
(PST)"
Date: 25 Jan 1999 09:55:07 -0800
Message-ID: <yl7lubutj8.fsf at windlord.stanford.edu>
Lines: 24
X-Mailer: Gnus v5.4.66/Emacs 19.34

Brian Kantor <brian at karoshi.ucsd.edu> writes:

> I also question whether there is any great win in having multiple feeds
> open between peers.  Assuming that the programs at both ends are running
> at full disk system saturation, there should be no advantage.

I have several peers where the first saturation point reached in a feed is
network latency, not disk, and opening multiple connections provides a
noticeable increase in throughput (the feed doesn't keep up without
multiple channels and does keep up with them).

> An exception to this is when the receiving system has its articles
> distributed on different disk systems by newsgroup or group type, and
> the feed splits its sending in the same manner.  Because there is some
> probability that this will occur unintentionally in a pairing which has
> not been explicitly set up this way, multiple overlapping feed channels
> can show some improvement over a single channel, but not consistently.

I think it's somewhat unlikely that multiple channels defined as carrying
the exact same traffic (which is the case here) are going to see that
different of distribution across disk systems.

--
Russ Allbery (rra at stanford.edu)         <URL:http://www.eyrie.org/~eagle/>