lkml.org 
[lkml]   [2009]   [Oct]   [26]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: [patch,rfc] cfq: merge cooperating cfq_queues
Date
Corrado Zoccolo <czoccolo@gmail.com> writes:

> Hi
> On Thu, Oct 22, 2009 at 2:09 AM, Jeff Moyer <jmoyer@redhat.com> wrote:
>> I think it's wrong to call the userspace programs broken.  They worked
>> fine when CFQ was quantum based, and they work well with noop and
>> deadline.
>
> So they didn't work well with anticipatory, that was the default from
> 2.6.0 to 2.6.17, and with CFQ with time slices, that was the default
> from 2.6.18 up to now. I think enough time has passed to start fixing
> those programs.

I actually didn't test anticipatory, so I'm not sure about that one.

> I think fixing nfsd at least for TCP should be easy. In TCP case, each
> client has a private thread pool, so you can just share the I/O
> context once, when creating those threads, and forget it.

I don't think it's a thread pool per client. Where did you get that
impression? Simply changing nfsd to use a single I/O context may be an
approachable solution to the problem. I'm not sure if it's optimal, but
it has to be better than what we have today.

> For the UDP case, would just reducing idle window fix the problem? Or
> the problem is not really the idling, but the bad I/O pattern?

I think the two cases can be handled the same way. I'll look into it if
time permits.

Cheers,
Jeff
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2009-10-26 16:09    [W:0.179 / U:0.364 seconds]
©2003-2011 Jasper Spaans. Advertise on this site