Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.ecn.uoknor.edu!feed1.news.erols.com!howland.erols.net!panix!news.panix.com!not-for-mail
From: eravin@panix.com (Ed Ravin)
Newsgroups: comp.security.firewalls,comp.unix.bsd.bsdi.misc
Subject: Re: Gauntlet/BSD "transitions" every hour
Date: 19 Apr 1997 20:12:03 -0400
Organization: All Watched Over by Machines of Loving Grace
Lines: 37
Message-ID: <5jbn0j$a4r@panix.com>
References: <5j8nrf$5dh@news4.digex.net>
NNTP-Posting-Host: panix.com
X-Newsposter: Pnews 4.0-test50 (13 Dec 96)
Xref: euryale.cc.adfa.oz.au comp.security.firewalls:7359 comp.unix.bsd.bsdi.misc:6677
From somewhere in cyberspace, tomk@access.digex.net (Tom Krotchko) said:
>
>I've got a TIS Gauntlet v3.2 installed on an Intel/BSDI system. It uses token ring
>cards (although I'm not sure that's relevant). Its a pentium with 16M of RAM.
>
>Every hour at about 17 minutes after the hour (EVERY hour, mind you),
>connectivity goes away on this machine. You can ping it, but you can't
>telnet, ftp, or http through it. At about 21 minutes after the hour,
>connectivity resumes. Its just like clockwork.
Is your Gauntlet installed at an institution that sends timing signals over
the AC line to synchronize mechanical clocks? If so, perhaps the BSDI
kernel thinks it is detecting a power failure. Can you switch the Gauntlet's
power to a UPS for testing?
>Weirder still, I just discovered today while trying to diagnose with
> systat -netstat
>
>That just after the connectivity comes back, I get the message at the bottom
>of the display:
> "Kernel In State of Transition"
Send those messages to BSDI support, or better yet, to TIS if you have
purchased Gauntlet support (it might be free for the first 30 days,
I forget the details).
Also, have you put a sniffer on either ring when the problem happens?
Can other hosts on either local ring communicate with each other?
Are your token rings plugged into smart token-ring MAU's that report
errors and de-insertions on the ring? You should investigate these
issues first to see whether the problem is in the Intel box or on your
network.
--
Ed Ravin |If I were not a little mad and generally silly
eravin@panix.com|I should give you my advice upon the subject, willy-nilly;
|I should show you in a moment how to grapple with the question
|And you'd really be astonished at the force of my suggestion.