[763] in resnet

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

Re: 3com switches' "all lights on" mode

daemon@ATHENA.MIT.EDU (Martin Hamilton)
Mon Feb 4 06:40:35 2002

Message-ID:  <E16XhR9-0001hI-00@gadget.lut.ac.uk>
Date:         Mon, 4 Feb 2002 11:37:27 +0000
Reply-To: Resnet Forum <RESNET-L@listserv.nd.edu>
From: Martin Hamilton <martin@NET.LUT.AC.UK>
To: RESNET-L@listserv.nd.edu
In-Reply-To:  Message from Rick Coloccia <coloccia@geneseo.edu> of "Fri, 01 Feb
              2002 15:15:27 EST."
              <5.1.0.14.2.20020201150548.03300878@imap.geneseo.edu>

Rick Coloccia writes:

| I've had several 3com switch 3300s (all varieties [plain 3300s, XM, MM, TM,
| SM]) fail in the exact way you explain (all the lights come one, one may be
| blinking.)

That's interesting - we used various flavours of 3300 quite
extensively without experiencing this problem, e.g. one of them was
the DMZ on our Internet link, and another was the core of our central
server LAN.

Dare I ask if anyone has come across it on the newer 3com switches,
e.g. 4400 ?

| In every case, I've been able to get 3com to replace the dead
| switches.  3300s are still made and supported, though.

In the UK at least, they will replace 610s if you return them via your
original reseller.  This presupposes you kept an inventory of which
switches came from which resller, of course :-)

| Sorry I don't have much to help you with, but it makes you feel any better,
| you aren't the only location having a hard time with 3com gear...

I was curious about the "wibbling" aspect - only hit on the "solution"
(tweaking uplink port settings) by accident, as I was sitting at home
one rainy weekend wondering if there was some way I could avoid having
to go over and restart a number of switches which had failed.

Suspect a lot of people are seeing their switches die, and assuming
they have to be visited and power cycled.

The fact that it's possible to recreate this failure mode suggests
some duff coding in the firmware, though perhaps the problem actually
runs as deep as the hardware.  This might explain why there has been
no fix forthcoming in the 2.6x firmware revisions.

Thanks again for any thoughts...

Cheers,

Martin

___________________________________________________
You are subscribed to the ResNet-L mailing list.

To subscribe, unsubscribe or search the archives,
go to http://LISTSERV.ND.EDU/archives/resnet-l.html
___________________________________________________

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