[Bioclusters] job queueing systems for ibm blade center

Rayson Ho bioclusters@bioinformatics.org
Sun, 26 Sep 2004 07:54:51 -0700 (PDT)

If you are talking about the x86 version of the BladeCenter, then the
pre-compiled version of SGE will run smoothly on your machines.

If you have the PPC version, then you will need to download the SGE
source tarball from the project homepage and compile it on your
machine. I use SGE 5.3p4 on a cluster of Power4s and Power3s running
Linux (in fact, SUSE and RHEL), and SGE works just fine there.

If you go with SGE 5.3, make sure you know how to turn on the SGEEE
mode, which allows you to setup projects and more advanced scheduling
features. For SGE 6.0, SGEEE mode is enabled by default, and you will
not see the term in the documentation.

SGE supports more platforms than TORQUE, but if you only use Linux,
then both of them work fine.

Since SGE/TORQUE are free, you can always try them. If neither of them
work, then you can go for LSF, but it is not free and it supports less
platforms than SGE.


--- Hershel Safer <hsafer@alum.mit.edu> wrote:
> Hi folks,
> I'd be interested to hear from folks who are using IBM Blade Center 
> clusters about good and bad experiences you've had with various job 
> queueing systems. I'm especially interested in SGE, LSF, and TORQUE.
> Our blades will be available to anybody in the university doing 
> bioinformatics or other biological computer analysis; we're not doing
> massive pipeline globally, though some researchers might do some for 
> specific projects. We're looking for a system that will queue and 
> distribute the jobs across the machines in the cluster.
> Thanks,
> Hershel

Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!