kernel panics and nvram issues

• Subject: kernel panics and nvram issues
• From: David Osguthorpe <email@hidden>
• Date: Tue, 8 Feb 2005 15:49:04 -0700
• Delivered-to: email@hidden
• Delivered-to: email@hidden
• User-agent: Mutt/1.4.2.1i

Hi,

I am having problems setting up G5 cluster nodes to dump core
to the head node (this relates to my previous issue with kernel
panics occurring infrequently and possible interactions with hwmond)

Ive setup the head node with appropriate xinetd.d entry to recieve
the dumps and on a test machine using Apples InstantPanic everything
works fine.

The problem is that although I have set the nvram boot-args variable
on all the nodes with the appropriate _panicd_ip debug flags which
worked on the test machine - at least running the nvram command
on the slave nodes claims this to be set up - most of the times
instead of getting a core dump Im getting a panic log - not only
that after the machine reboots the boot-args variable is reset
to data that it had before updating with _panicd_ip etc. 

Another twist is that recently a circuit breaker failed cutting power
instantly to a set of nodes - when power was restored and these nodes
rebooted again the boot-args variable was reset and no longer contained
the _panicd_ip data

Any ideas why nvram is not non-volatile?
Is there any other command that enforces updating the nvram hardware?

Thanks

David
 _______________________________________________
Do not post admin requests to the list. They will be ignored.
Darwin-dev mailing list      (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/darwin-dev/email@hidden

This email sent to email@hidden

• Follow-Ups:
  • Re: kernel panics and nvram issues
    • From: Shantonu Sen <email@hidden>

• Prev by Date: APSL 2.0 vs. APSL 1.x
• Next by Date: Re: kernel panics and nvram issues
• Previous by thread: Re: APSL 2.0 vs. APSL 1.x
• Next by thread: Re: kernel panics and nvram issues
• Index(es):
  • Date
  • Thread