2012-08-17 14:27:41

by Suresh Jayaraman

Subject: Slowing down state manager after unhandled error

Hi Trond and Chuck,

Recently, we have had a report on production SLES system where the
system's CPU utilization went very high due to the state manager
constantly sending huge number of requests every second due to an
unhandled error. The following patch from Chuck

https://patchwork.kernel.org/patch/1185421/

mitigated this issue. Would be interested to know whether this patch
queued up already or you see any problem with the approach.


Thanks
Suresh

2012-08-17 15:06:50

by Chuck Lever III

Subject: Re: Slowing down state manager after unhandled error

Hi-

On Aug 17, 2012, at 10:27 AM, Suresh Jayaraman wrote:

> Hi Trond and Chuck,
>
> Recently, we have had a report on production SLES system where the
> system's CPU utilization went very high due to the state manager
> constantly sending huge number of requests every second due to an
> unhandled error. The following patch from Chuck
>
> https://patchwork.kernel.org/patch/1185421/
>
> mitigated this issue. Would be interested to know whether this patch
> queued up already or you see any problem with the approach.

It got dropped for 3.6 simply because it was at the end of a patch series that was truncated in the middle. I'm submitting it again for 3.7.

--
Chuck Lever
chuck[dot]lever[at]oracle[dot]com