The last few weeks, we've been getting this kernel panic on one machine or another out of a cluster of 15, about one every four days. The machines are all running Ubuntu 10.04 LTS with Erlang R13B03; the machines have dual quad-core hyper-threaded Xeon E5520 CPUs.

The crash stack looks like the attached picture (which we had our co-lo provider send us from the crashed console):

Stack dump of kernel crash in load_balance_newidle called from thread_return

The linux version is:

Linux AF001783 2.6.32-28-generic #55-Ubuntu SMP Mon Jan 10 23:42:43 UTC 2011 x86_64 GNU/Linux

The weird thing is, these hosts have been running without this problem for a year before now, and the load profile is not significantly different from before. And it's not just a single host, in which case I'd suspect bad hardware.

  • You'll need to post more info than that. Is this reproducible? What are you doing right before the crash happens? Do you have the crashdump package installed? If so post the dump (to pastebin or something) and that will yield some more info. It's in /var/log/dump btw. – Zeroedout Sep 15 '11 at 6:21
  • Well, if this doesn't get resolved soon, feel free to upgrade your kernel (latest stable is reported to be 3.0.4). This may be a bug that has been fixed. – new123456 Sep 15 '11 at 11:06
  • Thanks for your comment, but you didn't understand the question. We're running Ubuntu 10.04 LTS, for the reason that it's Long Term Support. Others can live on the bleeding edge ;-) – Jon Watte Dec 20 '11 at 6:50

Browse other questions tagged or ask your own question.