lkml.org 
[lkml]   [2009]   [Apr]   [6]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 0/8][RFC] IO latency/throughput fixes
On Mon, Apr 06 2009, Linus Torvalds wrote:
>
>
> On Mon, 6 Apr 2009, Jens Axboe wrote:
> >
> > > Jens - I can just apply this queue (I want to test it out anyway), or if
> > > you prefer I can pull from you. Just tell me.
> >
> > Whatever you want, if you want to pull instead of manually applying,
> > it's:
> >
> > git://git.kernel.dk/linux-2.6-block.git blk-latency
>
> Ok, I applied them for testing anyway, so I think I'll just keep the
> series I have in my tree. I'm keeping my nasty "dd+sync" going in the

No problem, the end result should be identical :-)

> background, just to verify that things really feel better, but I'm already
> convinced this is a winner. Especially with the request size limiter, I
> can really read email _almost_ as if nothing else was going on on the
> machine.
>
> With that request size limiting, my fsync pauses tend to be in the 250ms
> range, with a one 0.75s outlier in the last few minutes. I can definitely
> feel that quarter-second thing, and the .75s pause was a real stutter, but
> boy what a difference. I don't get the uncontrollable urge to kill that
> nasty background writer any more.

If we can make the dd writer almost unnoticable, then I think we've come
a long way already. That really is a nasty workload.

--
Jens Axboe



\
 
 \ /
  Last update: 2009-04-06 19:03    [W:0.111 / U:0.400 seconds]
©2003-2014 Jasper Spaans. Advertise on this site