[Bioclusters] Sequest on Linux

Rayson Ho raysonlogin at yahoo.com
Wed Aug 17 14:10:09 EDT 2005


Does anyone know how PVM is used within Sequent?? If PVM is only used
as a task distribution mechanism, then we can simply submit serial
version of the jobs to the batch system.

Using a batch system (like SGE or LSF) provides more benefits, like job
control, load balancing, master node failover, accounting.

Tight SGE/PVM integration:
http://gridengine.sunsource.net/howto/pvm-integration/pvm-integration.html

AFAIK, LSF does not do accounting for PVM tasks (LSF parallel handles
MPI jobs, anyone knows if it also handles PVM??):

"Important: LSF does not time pvm jobs properly. The time reported by
lsf will only account for the master process, not the spawned slaves."

 www.ncsa.uiuc.edu/UserInfo/Resources/Hardware/CommonFAQ/SGIMPI.html

Rayson


--- "Brodie, Kent" <brodie at mcw.edu> wrote:
> For integration, my understanding is that A Sequest-installed series
> of
> systems can co-exist with other job scheduler environments on the
> same
> cluster.    As I mentioned earlier, RSH playes a huge part in keeping
> Sequest talking.    The technical communication between the nodes
> just
> really isn't that complicated, and my assumption is something like
> LSF/SGE/PBS/etc could peacefully co-exist.
> 
> If you have further scientific-like queries, Simon will tackle those,
> and I'll be happy to address any other sysadmin-like questions you
> may
> have.



__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


More information about the Bioclusters mailing list