[GE users] cputime for parallel jobs does not reflect the number of cpus used

Rayson Ho rayrayson at gmail.com
Thu Mar 2 18:56:18 GMT 2006


    [ The following text is in the "ISO-8859-1" character set. ]
    [ Your display is set for the "ISO-8859-10" character set.  ]
    [ Some special characters may be displayed incorrectly. ]

That's why you are not getting the CPU time accounted for the slave
MPI tasks... If tight integration is working correctly, you should see
that all the slave MPI tasks being children of the sge_shepherd.

You can search the list archive for replies from Reuti... he has
helped a lot of people on those parallel job issues...

Rayson



On 3/2/06, Nicolas Bock <nbock at lanl.gov> wrote:
> > When tight integration is on, and you logon to the nodes that
> > have the slave MPI tasks, do you see the MPI slave tasks started
> > by the SGE shepherd?
>
> Thanks to Gerd for clarifying that. So the answer then is no, they
> are not started by the shepherd.
>
> nick
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
> For additional commands, e-mail: users-help at gridengine.sunsource.net
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net




More information about the gridengine-users mailing list