[Bioclusters] software raid / ISMB rollcall / Platform question(s)

[Joe Landman]

On Wed, 2002-07-24 at 10:57, Chris Dagdigian wrote:
> 
> Hi folks,
> 
> Random bits to throw out...
> 
> (1) Anyone have any good methodologies to share regarding maximizing 
> linux-on-intel file I/O with cheap ATA drives and linux software RAID? 

But of course.  Depends upon your usage pattern. I have been working on
a writeup of this.  Email me offline if you need details before I post
it.

> We've been seeing some amazing numbers on a couple of prototype cluster 
> compute nodes using pairs of 80gig ATA drives controlled by a Promise 
> ATA card running reiserfs on top of software RAID0. The tests are not 
> complete yet but the numbers appear to be better than what we can 
> squeeze from (a) scsi drives, (b) a direct SAN connection and (c) a 
> single 100Tx connection to a Netapp F840 NAS filer.  We'll eventually 
> post what we find to be our 'best config' once our tests are done. We 
> also need to get numbers from the Netapp when we have a gigE link.

Depends upon many things.  Local drives on high speed interconnects are
almost always faster* than remote drives over lower speed interconnects
for large block sequential IO.

(* I have as of yet to hear of a case contrary to this, but I will keep
writing it as "almost always" as I am sure there is a pathological case
out there waiting to do damage to the word "always")


-- 
Joe Landman,
email: landman@scientificappliance.com
web  : http://scientificappliance.com