On 10/04/2011 06:08 PM, Florian Philipp wrote:
Am 04.10.2011 17:54, schrieb Florian Philipp:
Am 03.10.2011 12:50, schrieb carlopmart:
Hi all,
I have installed two SL6.1 (fully patched) hosts with cman+rgmanager
suite for HA services. When both nodes starts, this error is displayed:
"NOHZ: local_softirq_pending 08"
What does it means?? I have another pair of RHEL6.1 hosts with
cman+rgmanager and this error doesn't appears ...
All hosts are installed under ESXi 5.
Thanks
I've traced the piece of kernel code which causes this warning through
various revisions. It seems to be an old unresolved kernel bug triggered
on various machines.
Originally it was introduced as a bug assertion in [1]. Then it became a
warning in [2] in order to provide better debugging. Finally, in [3] it
was rate limited to stop it from filling up the kernel logs.
I suggest you either look for discussions of this bug on the linux
kernel mailing list or contact Thomas Gleixner (who provided these
patches) directly. His email address is listed in those commits.
[1]
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=79bf2bb335b85db25d27421c798595a2fa2a0e82
[2]
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=bc5393a6c9c0e70b4b43fb2fb63e3315e9a15c8f
[3]
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=352823160613b65fdaa558be486720a71f75ed86
Regards,
Florian Philipp
FWIW, check out this discussion and patch. It seems it was never applied
to the main sources.
https://lkml.org/lkml/2007/5/22/35
Many thanks Florian. Error seems to be displayed when vmxnet3 is used.
If I use another one like e1000, warning is not displayed. Strange ...,
--
CL Martinez
carlopmart {at} gmail {d0t} com
