[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Nirvana Queue Question
From: |
Jeffrey Oldham |
Subject: |
Nirvana Queue Question |
Date: |
Tue, 16 Oct 2001 16:01:13 -0700 |
I was trying to execute
bsub -K -q shared -n 49 'mpirun -np 49 ./SGI64KCC/Doof2d --sim-params 2800 0 1
--num-patches 7 --run-impls 5 --pooma-stats --pooma-nocompress -mpi >
results_mpi_5_p49_400_7' && pprof -m >> field_mpi_5_p49_400_7
The job got stuck for over an hour. Do you think this is user error
or did the job allocator refuse to allocate the job because every
possible 49-processor chunk already had something running? Note the
latter does not make since because there are more shared jobs running
than the 112 nodes running them. Also note I experienced no such
problem yesterday with similar jobs up to 36 processors.
Thanks for the information,
Jeffrey D. Oldham
address@hidden
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- Nirvana Queue Question,
Jeffrey Oldham <=