Changeset 7623


Ignore:
Timestamp:
Feb 12, 2010, 5:30:54 PM (9 years ago)
Author:
ole
Message:

Comment on parallel usage and added reporting of which nodes processes are using

File:
1 edited

Legend:

Unmodified
Added
Removed
  • anuga_validation/okushiri_2005/run_okushiri_parallel.py

    r5848 r7623  
    1919mesh before running this script.
    2020
     21To run on e.g. 8 processors
     22mpirun -machinefile ~/.machines_tornado -np 8 python run_okushiri_parallel.py
     23
     24where the machine file has the format
     25
     26compute-0-19 cpu=4
     27compute-0-18 cpu=4
     28compute-0-17 cpu=4
     29compute-0-16 cpu=4
     30compute-0-15 cpu=4
     31compute-0-14 cpu=4
     32compute-0-13 cpu=4
     33compute-0-12 cpu=4
     34compute-0-11 cpu=4
     35compute-0-10 cpu=4
     36compute-0-9 cpu=4
     37compute-0-8 cpu=4
     38compute-0-7 cpu=4
     39compute-0-6 cpu=4
     40compute-0-5 cpu=4
     41compute-0-4 cpu=4
     42compute-0-3 cpu=4
     43compute-0-2 cpu=4
     44compute-0-1 cpu=4
     45compute-0-0 cpu=4
     46tornado cpu=4
     47
     48See mpirun manual for more details
    2149"""
    2250
     
    3260from anuga.abstract_2d_finite_volumes.util import file_function
    3361
    34 from anuga_parallel.parallel_api import myid, numprocs, distribute   
     62from anuga_parallel.parallel_api import myid, numprocs, distribute, processor_name   
    3563
    3664import project
     65
     66print 'Pypar running process %d of %d on node "%s"' % (myid, numprocs, processor_name)
    3767
    3868use_cache = True
Note: See TracChangeset for help on using the changeset viewer.