[HTCondor-users] job folders left on cluster

["Ryter, Derek" <dryter@xxxxxxxx>]

This doesn't happen consistently, but at times a server will be completely filled with folders created by jobs in the condor\execute folder. I'm running pest++. Is it because of how a run terminates, or something missing in the submit file? Any ideas? Below is the submit file. Thanks.

#Mode under which Condor is running

universe = vanilla

#Log location

log = log/master_$(Cluster).log

output = log/worker_$(Cluster)_$(Process).out

error = log/worker_$(Cluster)_$(Process).err

#Want to return info to logs

stream_output = true

stream_error = true

#Run executable

executable = worker.bat

#Args ########################################################################

arguments = 136.177.176.31 4044 new_rch_rch.pst

requirements = ((Target.OpSys=="WINDOWS") &&Â (Target.Arch=="X86_64"))

#In Mb

request_memory =1000

#Choose the fastest machines

rank = MIPS

#YES = Transferring data

should_transfer_files = YES

when_to_transfer_output = ON_EXIT_OR_EVICT

transfer_input_files = unzip.exe, data.zip, worker.bat, python2.7.14.zip

queue 180

Cheers,

Derek Ryter, Ph.D. PG

USGS Oklahoma Water Science Center

Remote address:

86 Aspen Dr.Â

Durango, CO 81301

(402) 613-3323 (cell)

USGS Profile