Re: Ext2 directory index: ALS paper and benchmarks

Daniel Phillips (phillips@bonn-fries.net)
Thu, 6 Dec 2001 04:54:31 +0100


Hi Hans,

On December 6, 2001 04:41 am, you wrote:
> I can't comment on your benchmarks because I was on the way to bed when
> I read this. I am sure though that you and Stephen are doing your usual
> good programming.
>
> ReiserFS is an Htree by your definition in your paper, yes?

You've got a hash-keyed b*tree over there. The htree is fixed depth.

> Daniel Phillips wrote:
> >So it seems that for realistic cases, ext2+htree outperforms reiserfs
> >quite dramatically. (Are you reading, Hans? Fighting words... ;-)
>
> Have you ever seen an application that creates millions of files create
> them in random order?

We haven't seen an application create millions of files yet. However, the
effects I'm describing are readily apparent at much smaller numbers.

> Almost always there is some non-randomness in the
> order, and our newer hash functions are pretty good at preserving it.
> Applications that create millions of files are usually willing to play
> nice for an order of magnitude performance gain also.....

To be fair, I should rerun the tests with your linear-congruential hash, I'll
try to get time for that.

--
Daniel
-
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/