Unfortunately I was not yet able to pinpoint exactly the code that is
wrong. But I think I know a way that can be used as a work-around until
I find out the real fix. It looks very much like the problem is related
to the autogroup feature which will automatically put tasks in separate
task groups based on the session ids. The ohai process is doing this a
lot, so I think compared to other work-loads there is a high volume of
coming and going task groups, which might be the problem.

At the cost of loosing the fairer scheduling which isolation of high cpu using 
tasks in their own group brings, it is possible to disable the autogroup 
feature by either using "noautogroup" as pvgrub kernel argument, or changing
/proc/sys/kernel/sched_autogroup_enabled to 0. I guess that can be a sysctl 
entry, too. That way the ohai reproducer seems to survive, so production 
systems won't crash all the time (hopefully).

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/999755

Title:
  Kernel crash in rb_next doing ohai loops

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/999755/+subscriptions

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to