[vortex-bug] 3c905CX-TXM - One machine now works, one machine doesn't

Garnet Chaney gchaney@webmill.com
Thu, 7 Dec 2000 22:54:14 -0800


Bogdan,

I feel really silly. A new cable fixed the problem!  I now remember why I
accidentally took a bad cable with me when
I installed the machine.  Thanks for the hint.

- Garnet

-----Original Message-----
From: Bogdan Costescu [mailto:Bogdan.Costescu@IWR.Uni-Heidelberg.De]
Sent: Tuesday, December 05, 2000 5:02 AM
To: Garnet Chaney
Cc: vortex-bug@scyld.com
Subject: RE: [vortex-bug] 3c905CX-TXM - One machine now works, one
machine doesn't


On Mon, 4 Dec 2000, Garnet Chaney wrote:

>Working machine:
>  Station address set to 00:01:03:27:08:87.
>  Manufacture date (MM/DD/YYYY) 10/12/2000, division H, product HN.
>Nonworking machine:
>  Station address set to 00:01:03:27:08:a7.
>  Manufacture date (MM/DD/YYYY) 10/12/2000, division H, product HN.

They are really very close...

>   Hurricane format checksum is correct (0xb0 vs. 0xb0).

This is a bit strange. Usually, the C cards fail to have a correct
checksum; are you sure that this card (the non-working one) hasn't been
subject to the DOS 3Com config utility or something similar ? This might
influence the media selection as well.

> I am starting to wonder if maybe the card was jolted or something when I
> moved the machine to my ISP. Could the card still be responding to these
> kinds of diagnostics, even without being squarely in the slot?

Not really... If it's not firmly inserted in the slot, you would get
garbage out of it, while here we have coherent data.

> I will try a different switch tommorrow, or try switching ports on the
> current switch.

Try switching the cables too...

Sincerely,

Bogdan Costescu

IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
E-mail: Bogdan.Costescu@IWR.Uni-Heidelberg.De