Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 24 Jun 2002 21:46:18 -0700 (PDT)
From:      Lamont Granquist <lamont@scriptkiddie.org>
To:        "Morse, Richard E." <REMORSE@PARTNERS.ORG>
Cc:        freebsd-stable@freebsd.org
Subject:   RE: Did someone break TCP?
Message-ID:  <20020624214242.G20202-100000@coredump.scriptkiddie.org>
In-Reply-To: <375F68784081D511908A00508BE3BB1701EF1BEF@phsexch22.mgh.harvard.edu>

next in thread | previous in thread | raw e-mail | index | archive | help


On Mon, 24 Jun 2002, Morse, Richard E. wrote:
> Lamont Granquist [mailto:lamont@scriptkiddie.org] wrote:
> > I've got an IRC server with a few days of uptime now after
> > upgrading to
> > both FBSD 4.6 and ircd-hybrid 6.3 out of the ports tree.
> > Tonight I found
> > that I couldn't login, TCP wasn't completing the 3-way
> > handshake (no RST,
> > no SYN|ACK).  Stopping the service and restarting it fixed
> > the problem.
> > At this point it looks like TCP freaked out.
> >
> > I tried multiple different ways to login to the ircd, both
> > remotely, from
> > the local box and from behind the machine that it acts as a
> > firewall for.
> > All the users reported a 3+ hour outage prior to me finding
> > the problem.
> > There was nothing that I was doing to the box at the time.
> > This has been
> > previously stable hardware.
>
> Hi!  This sounds like a problem I'm having on a FreeBSD 4.5-p6 (actually,
> any of the patch levels) box.  After a few days, it just drops off the
> network.  If I try to ping anything, I get an error "no buffer space
> available" (I think -- I'm remembering this off the top of my head).  I've
> written about this a number of times, and no one has been able to suggest a
> solution.  I've discovered that running tcpdump will force the computer to
> reconnect to the network, so the best that I've been able to do (which cures
> the symptoms, but not the problem) is to run 'tcpdump -c 10' via cron every
> two hours or so...
>
> If you do figure out what causes this, _please_ let me know.  I would
> _really_ like to see this fixed...

No, that isn't the same issue.  I could still ssh into the box, and
therefore don't believe it was a driver issue.  It seemed to be that
something in the data structures associated with that one socket got
corrupted?


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020624214242.G20202-100000>