[Beowulf] OS for 64 bit AMD

Joe Landman landman at scalableinformatics.com
Wed Apr 6 07:43:08 PDT 2005

SSI is hard to do right.  You really need help at the bios level of each 
node (at boot up they join a master graph/data structure that abstractly 
represents the hardware).  DSM's reside atop this.  Linux can DSM just 
fine (look at the Altix).  Scyld doesn't do DSM but they do SSI.

So far on the Debian side, I have been fairly impressed with Ubuntu in 
terms of packaging and ease of install.  If we were asked to build 
Debian or Debian based machines, this is likely what we would prefer to 
use.  It installed without a hitch on machines that gave other 
distributions migranes and worse.  Personally I like the idea of a 
knoppix variant as a core (minimal core OS installed per machine), and 
layered bits upon that.  I know there is a clusterknoppix and quantian.

I am sorry to hear about the broken drivers in FC-2.  FC-3 seemed to fix 
some of the borked drivers.  Lots of closed source ones also released 
around then to accomodate the changes to the kernel.  You might try FC-3 
as well.

If you find the distro you like, you might want to have a look at 
Warewulf atop that as a cluster meta-distro.


Tony Travis wrote:
> Robert G. Brown wrote:
>> [...]
>> Ah, but the PROPER and ORIGINAL meaning of axiom is "unprovable
>> assumption" upon which a system of logic and discourse is built.  The
>> final half-definition is the only one that is actually correct.
>> In fact, I've got a whole essay entitled "Axioms" on my personal
>> website.  Seriously.
> Hello, Robert.
> OK, I bow to your greater wisdom: I have now learned that axioms are:
> "...unprovable assumptions whose truth and falsehood cannot be assessed"
> [http://www.phy.duke.edu/~rgb/Philosophy/axioms.php]
> What struck me was that both Mark and Joe used the word "axiom" 
> appropriately according to at least one dictionary:
>     http://www.thefreedictionary.com/axiom
> I don't think their discussion was pointless: Actually, I found it 
> interesting and relevant to my own situation running openMosix on a 
> 64-node RH-9 based Beowulf cluster. I had a difficult time trying to 
> upgrade our cluster from RH-8.0 to FC-2 because the Adaptec Ultra160 
> drivers in FC-2 were broken, so I went to RH-9 instead.
> I'm now deciding where to go after RH-9 and I'm interested in other 
> people's experiences. I've already moved one cluster server to Debian 
> Sarge/testing, but I'm also evaluating the recent Progeny Debian 
> Developer Edition RC2 which includes a Debian port of Red Hat's 
> Anaconda. Have you any experience using Progeny Debian?
> I read that Progeny originally set out to build Debian SSI cluster 
> software, but had problems and abandoned the project. Is that true?
> Best wishes,
>     Tony.

Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web  : http://www.scalableinformatics.com
phone: +1 734 786 8423
fax  : +1 734 786 8452
cell : +1 734 612 4615

More information about the Beowulf mailing list