> On Wed, Jan 31 2001, Steve Lord wrote:
> > It is difficult to nail down the cause yet, but dbench has taken a
> > significant hit (40%). Single file (bonnie) is comparable with 2.4.0, the
> > dbench issue may be related to the throttling being placed on the
> > ll_rw_block interface by Jens' changes. The problem with working out
> > a bottleneck in dbench is it is such a random pile of stuff that
> > you really cannot point a finger at any one thing. I need to go
> > find some more deterministic measurements.
> >
> > The performance drop off is not limited to xfs though ext2 appears
> > to suffer as well.
>
> You can easily disable the throttling of locked buffers by just
> defining blk_started_io and blk_finished_io to do nothing. What
> disk are you using, and how much RAM in the machine?
This is a Segate ST39175LW - nothing to exciting. I am actually
now not sure if the throttling code really has too much to do with it.
In answer to your next message, yes dbench appears to be the only
thing I can find which really suffers, so I am not too worried.
Currently I am working my way though a deadlock I managed to introduce
with some other changes, then I will be pushing the 2.4.1 version
out onto the net.
Steve
>
> Also, is dbench the only bench affected?
>
> --
> Jens Axboe
|