[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] How to troubleshoot MPI job?

On Wed, Feb 16, 2005 at 11:51:23PM +0800, Nigel Teow wrote:
> My qns are,
> 1. How do I check which mpich Condor is using to run the job?

It's whatever MPICH you used to build your program with (ie whatever
mpicc you used)

> 2. What does Condor need from mpich to run the mpi job?

Nothing. We use 0.0% of MPICH at runtime. It's why you can't/don't need
to use ssh with the MPI universe. This is very different than every other
batch system out there, which is why it's so confusing for people.

> 3. Is there a way to monitor/observe how Condor runs the mpi job?

The shadow log on the submit machine, and the starter logs on the execute

> 4. How does Condor know which machines to use?

It matchmakes with the dedicated resources it has.