[45786] in Hotline Meeting

home help back first fref pref prev next nref lref last post

LOG # 125900

daemon@ATHENA.MIT.EDU (Louis A. Isgur)
Tue Nov 10 18:56:33 1998

To: hotline@MIT.EDU
Date: Tue, 10 Nov 1998 18:56:28 EST
From: "Louis A. Isgur" <larugsi@MIT.EDU>


------- Forwarded Message

Received: from SOUTH-STATION-ANNEX.MIT.EDU by po9.MIT.EDU (5.61/4.7) id AA12446; Tue, 10 Nov 98 18:45:31 EST
Received: from STINKURD.MIT.EDU by MIT.EDU with SMTP
	id AA03921; Tue, 10 Nov 98 18:45:19 EST
Received: by stinkurd.mit.edu (950413.SGI.8.6.12/4.7) id SAA26218; Tue, 10 Nov 1998 18:45:20 -0500
Received: from Messages.8.5.N.CUILIB.3.45.SNAP.NOT.LINKED.stinkurd.mit.edu.sgi.52
          via MS.5.6.stinkurd.mit.edu.sgi_52;
          Tue, 10 Nov 1998 18:45:20 -0500 (EST)
Message-Id: <QqGB0EoGgE6gBXmEJn@mit.edu>
Date: Tue, 10 Nov 1998 18:45:20 -0500 (EST)
From: Bill Cattey <wdc@MIT.EDU>
To: bmurphy@MIT.EDU, rbasch@MIT.EDU, ghudson@MIT.EDU, mbarker@MIT.EDU,
        network@MIT.EDU, ops@MIT.EDU, Alex T Prengel <alexp@MIT.EDU>
Subject: Further testing by wdc and larugsi
Cc: alexp@MIT.EDU, cluster-services@MIT.EDU, fuzzballs@MIT.EDU,
        nschmidt@MIT.EDU
In-Reply-To: <199811102246.RAA08499@zolak.mit.edu>
References: <199811102246.RAA08499@zolak.mit.edu>

Lou and I went to 4-035 and did some testing.

Tue Nov 10 17:53, it took me 2 and a half minutes to log into m4-035-13,
start fvwm and 3 xterms.  That's much slower than my home O2.

Lou started biosym on M4-035-12 while I did a ping of my home machine.
It took forever to show up.

The ping showed 33% packet loss.

When the ping quieted down I had  Lou run biosym from m4-035-11.  It
started up pretty quick.  (This was to confirm that starting biosym did
not hose the net.)

I did an
fs cachesize 1
fs cachesize -blocks 74826
to flush out the AFS cache on m4-035-12 to see if pumping bits in was
taking a long time.

While I did ping, I started biosym again.
The ping showed no packet loss.
biosym started VERY quickly.

My opinion, is that there is some intermittent thing going on with that
subnet that is causing massive packet loss, and that that is causing the
problem.

- -wdc

------- End of Forwarded Message


home help back first fref pref prev next nref lref last post