Re: Disk performance degradation

From: Stefano Rivoir (s.rivoir@gts.it)
Date: Wed Jul 30 2003 - 11:06:18 EST


Andrew Morton wrote:

> Stefano Rivoir <s.rivoir@gts.it> wrote:

>>Thanks for the hint. This seems to make things a little better, but I'm
>>still far away from 2.4 performances. I thought that anticipatory sched
>>could be part of the problem, and booting with elevator=deadline
>>does a little better... but using 2.4 is completely another thing.
>>By the way, -a 512 vs -a 8 is a kernel "change" or an hdpam one?
>
>
> What makes you think it is a disk performance problem at all? All we know
> is that KDE applications take longer to start up, yes?

> How much memory is in that machine? Can you run a `vmstat 1' trace during
> the "slow" operations?

I think I've got it. 2.4 fails to load DRI, so when X is up there is
memory available until the load of gnucash, the last operation. 2.6
loads dri and probably this eats too much too early, causing the
system to touch swap since the first operation after X startup. This
does not happen anymore disabling DRI in X.

Sorry for wasting your time :(

Thanks.

-- 
Stefano RIVOIR
GTS Srl

- 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/



This archive was generated by hypermail 2b29 : Thu Jul 31 2003 - 22:00:46 EST