File systems performance

[warlord at MIT.EDU (Derek Atkins)]

I think it all depends on the dataset..   Large or small files?  Large 
or small
directories?

.. and operations.. Creates?  Modifies?  Deletes?

The various FSes have different behaviors across this multi-dimensional 
matrix.

-derek

Quoting Jerry Feldman <gaf at blu.org>:

> I was in Houston for an HP/Intel Developer forum and was asked about file
> system performance. This was from a guy at a company who has a database
> product, and is interested in performance and not journaling. My top of the
> head answer was that ext2 would probably be the best because it does not
> have the journaling overhead, but I later checked some benchmarks, and
> found that ext2 did not always give the best performance.  My advice to him
> was to run their own benchmarks since they were more familiar with their
> product and the data metrics. What I'm looking for here is possibly some
> data you might have accumulated.
>
> (BTW: a number of the benchmarks show ext3 reasonably slow in comparison to
> JFS, ReiserFS, and XFS).
> --
> Jerry Feldman <gaf at blu.org>
> Boston Linux and Unix user group
> http://www.blu.org PGP key id:C5061EA9
> PGP Key fingerprint:053C 73EC 3AC1 5C44 3E14 9245 FB00 3ED5 C506 1EA9
> _______________________________________________
> Discuss mailing list
> Discuss at blu.org
> http://olduvai.blu.org/mailman/listinfo/discuss
>



-- 
       Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
       Member, MIT Student Information Processing Board  (SIPB)
       URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
       warlord at MIT.EDU                        PGP key available