Thank you to everyone who responded!
Everyone mentioned the battery.... some included additional information:
Werner Hahling
Dr. Udo Grabowski
Selden E. Ball, Jr.
Bob Lang
Kevin Dea
Peter Reynolds (who pointed out that INIT is sometimes needed and some problems that can show up due to switch settings)
Dr. Thomas P. Blinn
Alan Davis
(who both suggested using lan_config to set the values, which, as Dr. Blinn points out, make the SRM setting a moot point - as long as it's not a network boot that's involved)
I forgot to mention some details on the machine(s), but the information was all good either way. They are Raytheon M960 computers, based on the Alpha chip, and in our case, the settings are saved in flash, as a battery could not be permitted in the environment they will operate in.
As for the reason it happened, Alan Davis' mention of init completion leads me to suspect that the offending machine was powered off before the shutdown was complete.
For corrective measures, I have implemented the lan_config in the inet.local and all seems to be working for now.
Thanks to everyone!
-Patrick Norris
patrick.norris_at_trw.com
> -----Original Message-----
> From: Norris, Patrick [mailto:Patrick.Norris_at_trw.com]
> Sent: Monday, January 14, 2002 7:21 PM
> To: tru64-unix-managers_at_ornl.gov
> Subject: Self-Resetting firmware?
>
>
> Managers,
>
> We've seen, twice now, SRM settings on our Alpha's that have
> changed without any particular reason.
>
> It's not all the settings, for instance, this last time, it
> just dropped the ethernet speed down to 10MB, half.
>
> Admittedly, we bring these systems up and down in the lab
> more than most folks would, but I was wondering if anyone out
> there had seen this, and/or might have an explanation (that
> we can pass on to the powers-that-be -- it's causing no
> worries, considering the environment, among the folks working
> on the machines)?
>
> Thanks,
> -Patrick Norris
> patrick.norris_at_trw.com
>
Received on Thu Jan 17 2002 - 03:18:23 NZDT