Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 13 Jan 2022 08:26:19 -0800
From:      Gleb Smirnoff <glebius@freebsd.org>
To:        George Neville-Neil <gnn@neville-neil.com>
Cc:        net@freebsd.org
Subject:   Re: compressed TIME-WAIT to be decomissioned
Message-ID:  <YeBSq1GXTAa8CaN1@FreeBSD.org>
In-Reply-To: <C3A1E39F-B8A9-43D9-8813-A96227712B6F@neville-neil.com>
References:  <Yd8im/VkTU1zdvOi@FreeBSD.org> <C3A1E39F-B8A9-43D9-8813-A96227712B6F@neville-neil.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Jan 12, 2022 at 05:01:51PM -0500, George Neville-Neil wrote:
G> > * Who told that 2*MSL (60 seconds) is adequate time to keep TIME-WAIT?
G> >   In 71d2d5adfe1 I added some stats on usage of tcptw and experimented a bit
G> >   with lowering net.inet.tcp.msl. It appeared that lowering it down three
G> >   times doesn't have statistically significant effect on TIME-WAIT use stats.
G> >   This means that the already miniscule number of TIME-WAIT connection on a
G> >   modern HTTP server can be lowered 3 times more.  Feel free to lower
G> >   net.inet.tcp.msl and do your own measurements with
G> >   'netstat -sp tcp | grep TIME-WAIT'.  I'd be glad to see your results.
G> 
G> The origin of the 2*MSL is pretty old and from a different type of network, but, my understanding of your proposal is not a change to this value anyway, is that correct?  The removal of tcptw is a separate issue, if I understand you correctly.

Right. I'm not suggesting to change default MSL in FreeBSD. Just saying
that my experiments shown that it can be divided down by three with no
effects. And those people who are really concerned with amount of TIME-WAITs
hanging around can do that on their machines.

As you say "2*MSL is pretty old and from a different type of network", so
it probably can be reconsidered. But that's a separate thing to discuss.

-- 
Gleb Smirnoff



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?YeBSq1GXTAa8CaN1>