* wip-filestore
@ 2012-12-03 11:38 Sage Weil
2012-12-03 19:31 ` wip-filestore Samuel Just
0 siblings, 1 reply; 3+ messages in thread
From: Sage Weil @ 2012-12-03 11:38 UTC (permalink / raw)
To: sam.just; +Cc: ceph-devel
While playing with recovery on the plane I saw big latency spikes caused
by draining the filestore op queue. This patch makes it only wait for
dequeued ops (current handed to threads) and not anything else. Passed
the rados suite without problems, and some basic testing on my laptop with
lots of filestore op threads.
I remember you had a patch doing something similar a while back and
rejected it... but I forget what the issue was?
sage
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: wip-filestore
2012-12-03 11:38 wip-filestore Sage Weil
@ 2012-12-03 19:31 ` Samuel Just
2012-12-03 19:36 ` wip-filestore Mark Nelson
0 siblings, 1 reply; 3+ messages in thread
From: Samuel Just @ 2012-12-03 19:31 UTC (permalink / raw)
To: Sage Weil; +Cc: ceph-devel@vger.kernel.org
I rejected it only because I didn't manage to measure a performance
increase. The wip-filestore patches look fine though.
-Sam
On Mon, Dec 3, 2012 at 3:38 AM, Sage Weil <sage@inktank.com> wrote:
> While playing with recovery on the plane I saw big latency spikes caused
> by draining the filestore op queue. This patch makes it only wait for
> dequeued ops (current handed to threads) and not anything else. Passed
> the rados suite without problems, and some basic testing on my laptop with
> lots of filestore op threads.
>
> I remember you had a patch doing something similar a while back and
> rejected it... but I forget what the issue was?
>
> sage
>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: wip-filestore
2012-12-03 19:31 ` wip-filestore Samuel Just
@ 2012-12-03 19:36 ` Mark Nelson
0 siblings, 0 replies; 3+ messages in thread
From: Mark Nelson @ 2012-12-03 19:36 UTC (permalink / raw)
To: Samuel Just; +Cc: Sage Weil, ceph-devel@vger.kernel.org
Are we thinking about including this in 0.56?
Mark
On 12/03/2012 01:31 PM, Samuel Just wrote:
> I rejected it only because I didn't manage to measure a performance
> increase. The wip-filestore patches look fine though.
> -Sam
>
> On Mon, Dec 3, 2012 at 3:38 AM, Sage Weil <sage@inktank.com> wrote:
>> While playing with recovery on the plane I saw big latency spikes caused
>> by draining the filestore op queue. This patch makes it only wait for
>> dequeued ops (current handed to threads) and not anything else. Passed
>> the rados suite without problems, and some basic testing on my laptop with
>> lots of filestore op threads.
>>
>> I remember you had a patch doing something similar a while back and
>> rejected it... but I forget what the issue was?
>>
>> sage
>>
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2012-12-03 19:36 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-12-03 11:38 wip-filestore Sage Weil
2012-12-03 19:31 ` wip-filestore Samuel Just
2012-12-03 19:36 ` wip-filestore Mark Nelson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox