Date: Tue, 16 Mar 2004 15:04:07 +0300 From: Zherdev Anatoly <tolyar@mx.ru> To: Andre Oppermann <andre@freebsd.org> Cc: freebsd-net@freebsd.org Subject: Re: Problem with closing tcp session between cisco and freebsd Message-ID: <20040316150407.5d9eb9f4@dwarf.demos.su> In-Reply-To: <4056D84C.514EC45C@freebsd.org> References: <20040316125335.5f64cac5@dwarf.demos.su> <4056CFEA.6A231A9@freebsd.org> <20040316131256.015a082d@dwarf.demos.su> <4056D84C.514EC45C@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 16 Mar 2004 11:34:52 +0100 Andre Oppermann <andre@freebsd.org> wrote: > Do you know when it was working correctly the last time? I can't say closely because i first time look that situation on tcp level. But i have rlogin sessions that was'nt closed but cisco send stop packet in RADIUS. This was in april. I can't see tcp session because monitoring tool kill this rlogin. But looks like that it is the same situation. > Do you have > another machine (4.9R or lower, or 5.2 or 5.C) to test against? On I have servers with newer FreeBSD but not older. After problem with FreeBSD-SA-04:04.tcp i upgrade all my servers. And no -CURRENT on it, only -STABLE. > > March 2 the commit of the TCP segment reassembly queue limiter has > been done but that does not change TCP processing otherwise in any > way. A month ago the TCP-MD5 stuff has been committed. A year ago > there have been some NewReno fixes. > > So no obvious suspect. Before digging deep into the code its better > to have some more surrounding information. This situation not stable and may not be represented in short time. I have many rlogin sessions, but only in april was something similar. -- Zherdev Anatoly.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040316150407.5d9eb9f4>