When compiled with TM support, LAM uses "native" Torque support for
launching its MPI jobs. As such Torque is therefore aware of all the
MPI processes on all nodes, and can account for the CPU time used on
all nodes.
MPI's that do not support Torque's "native" launching support will use
rsh/ssh, and therefore Torque is not aware of all the MPI processes
launched on non-mother-superior nodes.
As such, LAM with TM support is reporting a more correct total CPU
usage number.
On Nov 5, 2007, at 1:50 AM, SCIPIONI Roberto wrote:
> Dear all,
>
> I compiled LAM MPI with torque using the option
>
> --with-boot-tm
>
> but now the qstat time seems to be multipled compared to the
>
> qstat -a one
>
> now qstat seems to give the total CPU time N CPU multiplied by the
> effective time, why ?
>
> Any way out ?
>
>
> Roberto S.
>
> ICYS, CLUSTER
> ICYS, NIMS
> Japan
> I can verify this behavior, but it doesn't happen all the time.
>
> Brock Palen
> Center for Advanced Computing
> brockp_at_[hidden]
> (734)936-1985
>
>
> On Nov 2, 2007, at 1:07 PM, Kamil Kisiel wrote:
>
>> Hello,
>>
>> Users of our cluster are experiencing some terminal issues when
>> using qsub -I. Console applications such as Vim do not resize when
>> the user resizes their terminal. Long commands in the shell wrap
>> back to the start of the line and overwrite characters instead of
>> continuing on to the next line.
>>
>> If the users ssh to the node (not using qsub -I), everything works
>> as expected.
>>
>> Has anyone else seen this issue?
>>
>> ____________
>> Kamil Kisiel
>> HPC Technician, Zymeworks Inc.
>> 201-1401 West Broadway,
>> Vancouver, BC, V6H 1H6, Canada
>> Tel: (604) 678-1388 ext. 35
>> Fax: (604) 737-7077
>> www.zymeworks.com
>>
>>
>>
>>
>>
>> Notice of Confidentiality: The information transmitted is intended
>> only for the person or entity to which it is addressed and may
>> contain confidential and/or privileged material. Any review, re-
>> transmission, dissemination or other use of or taking of any action
>> in reliance upon this information by persons or entities other than
>> the intended recipient is prohibited. If you received this in error
>> please contact the sender immediately by return electronic
>> transmission and then immediately delete this transmission
>> including all attachments without copying, distributing or
>> disclosing the same.
>> _______________________________________________
>> torqueusers mailing list
>> torqueusers_at_[hidden]
>> http://www.supercluster.org/mailman/listinfo/torqueusers
>
> _______________________________________________
> torqueusers mailing list
> torqueusers_at_[hidden]
> http://www.supercluster.org/mailman/listinfo/torqueusers
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
--
Jeff Squyres
Cisco Systems
|
