[Condor-users] MPICH1 jobs and Condor

["Mehdi Sheikhalishahi" <mehdi.alishahi@xxxxxxxxx>]

Hi,
 I configure requirements for this purpose(dedicated Resources and
dedicated Scheduler). I can successfully submit sleep and cat jobs in
parallel to Condor by Parallel Universe. Now  I want to submit a
MPICH1 job to Condor Pool. I use sample program of MPICH1 cpi and
define the following submit script.
---------------------------------------------------------------
[grid@Server Parallel]$ cat mpi1.sub
######################################
## Example submit description file
## for MPICH 1 MPI
## works with MPICH 1.2.4, 1.2.5 and 1.2.6
######################################
universe = parallel
executable = ./mp1script
arguments = ./cpi
machine_count = 2
output = outfile.$(NODE)
error = errfile.$(NODE)
log = logfile
#should_transfer_files = yes
#when_to_transfer_output = on_exit
#transfer_input_files = my_mpich_linked_executable
queue
---------------------------------------------------------------
I submit job by "condor_submit mpi1.sub" command, Then I see the
following errors in created output and error files.
---------------------------------------------------------------
[grid@Server Parallel]$ cat outfile.0
error 0 chirp putting identity keys back

[grid@Server Parallel]$ cat errfile.0
Can't connect to chirp server
---------------------------------------------------------------
What's the problem?
--
Best Regards,
S.Mehdi Sheikhalishahi,
Web: http://www.cse.shirazu.ac.ir/~alishahi/
Bye.