Re: elevator algorithm considered irrelevant

Rik van Riel (H.H.vanRiel@phys.uu.nl)
Wed, 18 Nov 1998 10:19:34 +0100 (CET)


On Tue, 17 Nov 1998, Mark Lord wrote:
> Larry McVoy wrote:
> > I wanted to mention that back around 1990 or so I spent a bunch of
> > time trying improve performance by playing with the disk sort routine
> > (the elevator alg) in SunOS. I futzed around for several weeks before
> > realizing that it didn't much matter what I did. That was the insight.
>
> Base assumptions may have changed somewhat since then..
> Drive speeds are now faster, partitions are much larger,
> Linux is not SunOS (async vs. sync I/O), ext2fs is not ufs.

Disks have gotten slower in comparison with the rest
of the system, so more expensive opimizations might
be worth it now.

> I once implemented "closest seek" and "2way-elevators" for
> the Linux IDE driver, and saw marked response improvements
> under heavy load. In practice, starvation of the edges was
> a non-issue, though I'm sure a pathological case could be
> conceived that might be noticeably malaffected.

I know what case: the case where people have installed Linux
on one partition with the swap partition a bit out of the way.

Since about 80% (maybe more) of all new installers seem to
choose this layout (usually with the swap partition at the
slower end of the disk:), I think we might want to keep this
situation in mind...

Conclusion: a one-way elevator is the way to go.

cheers,

Rik -- slowly getting used to dvorak kbd layout...
+-------------------------------------------------------------------+
| Linux memory management tour guide. H.H.vanRiel@phys.uu.nl |
| Scouting Vries cubscout leader. http://www.phys.uu.nl/~riel/ |
+-------------------------------------------------------------------+

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/