offtopic, was Re: AMI megaraid performance (was: RE: Dell release of Redhat 7.2)

Dirk Wetter dirkw at rentec.com
Fri Nov 9 15:53:01 CST 2001


On Fri, 9 Nov 2001, Arjan van de Ven wrote:

> On Fri, Nov 09, 2001 at 05:29:19PM -0500, Dirk Wetter wrote:
>
> [I don't want to have a flameware and I didn't intend to start one]

[ me neither ]

> I don't object to ANYONE using any kernel as base (assuming the work is open
> source etc); the thing I objected to in this case was using the SGI XFS
> kernel for performance measurement and then suggesting that this is the
> performance of the Red Hat kernel.

I saw the benchmarks, too. However the used fs should have definitely
an impact on the results, which somebody has to test with the same
benchmark. XFS uses synchronous writes/unlinks, whereas e.g. reiserfs
does asynchronous writes in addition it seems to have a large mem journal
log.

> Now the person who made that oversight doesn't work for SGI and I also don't
> blame him; it's a very easy mistake to make since the SGI XFS kernel (the
> RPM one, not the CVS one) has the same version number and all as the kernels
> Red Hat releases. The ease of this confusion was the cause of my worry.

ah...ok. I didn't know that. In fact: confusing.

cheers,
	~dirkw




More information about the Linux-PowerEdge mailing list