> >> I'm inclined to agree with Axel.
> >>
> >> You are proposing and trying to get advice on a configuration that does 
> >> not fit nicely with the current capability of torque and maui.
> >>
> >> There are lots of creative people reading who might give you good 
> >> suggestions (like virtualize or run multiple torque instances), but 
> >> perhaps you would get a more useful response if you were to describe the 
> >> problem you are trying to solve rather than the proposed solution you are 
> >> trying to implement.
> >>
> >> ie. What are you trying to achieve? - Why do you want to 'split CPUs on 
> >> each nodes between different queues'?
> > 
> > each node has GPU.  Jobs which use GPU use only one CPU in node. 
> > So I wish to make special queue for this kind of jobs and still use other 
> > CPUs for jobs which need only CPUs.
> > Of course we can directly ask the nodes in batch script but in this case 
> > idea of batch system on cluster with hundred nodes will be lost.
> 
> Wouldn't setting up a resource named GPU with total value 1 on each node
> solve your problem?
No, it does not solve. And I think you are right that scheduler should know how 
operate with this resource name.

> I'm not sure about Maui, but I have a patched version of Torque and FIFO
> scheduler that supports generic resources on nodes.
I tried torque  2.1.8 & 2.3.6, maui 3.26p21

Bye.


_______________________________________________
mauiusers mailing list
[email protected]
http://www.supercluster.org/mailman/listinfo/mauiusers

Reply via email to