|
Last change
on this file since 3315 was
3164,
checked in by jack, 19 years ago
|
|
Added code samples for batch scripts. Need to update results.tex to make it less confusing
|
|
File size:
892 bytes
|
| Rev | Line | |
|---|
| [3164] | 1 | #!/bin/bash |
|---|
| 2 | #PBS -l walltime=180 |
|---|
| 3 | #PBS -l nodes=2:ppn=2 |
|---|
| 4 | #PBS -l vmem=400mb |
|---|
| 5 | |
|---|
| 6 | ######################### |
|---|
| 7 | # The above #PBS directives do the following: |
|---|
| 8 | ######################### |
|---|
| 9 | # #PBS -l walltime=180 |
|---|
| 10 | # This job will not take longer than 180 seconds. |
|---|
| 11 | # |
|---|
| 12 | # #PBS -l nodes=2:ppn=2 |
|---|
| 13 | # This job requires 4 processors. Bogong has dual-processor nodes |
|---|
| 14 | # So 2 nodes * 2 processors-per-node (ppn) = 4 processors. |
|---|
| 15 | # |
|---|
| 16 | # #PBS -l vmem=400mb |
|---|
| 17 | # This job will not consume more than 400mb. |
|---|
| 18 | # |
|---|
| 19 | ######################### |
|---|
| 20 | |
|---|
| 21 | ######################### |
|---|
| 22 | # Run the MPI job. |
|---|
| 23 | # Note that the mpirun script specifies 4 processors, but 2 |
|---|
| 24 | # processors per node. Working directory is also set on the |
|---|
| 25 | # mpirun command line. |
|---|
| 26 | ######################### |
|---|
| 27 | cd /home/jack/ga/inundation |
|---|
| 28 | mpirun -machinefile $PBS_NODEFILE -np 4 -ppn 2 -wdir /home/jack/ga/inundation/parallel /home/jack/ga/inundation/parallel/run_parallel_sw_merimbula_metis.py |
|---|
Note: See
TracBrowser
for help on using the repository browser.
