Job data and executable are transferred directly from the submit node to the execute node. The central manager will not need space to store this data.
For a condor-c configuration, job data and executable will be transferred from submit node to submit node,
and job results will be held in the spool directory of the intermediate submit until the originating submit can fetch them. But even in this configuration the central manager still doesn't ever see the job data or executable.
I am looking at setting up our central manager on a CentOS virtual machine but need to make a decision about disk size for the image.
Do all queued job executables and input files get held at the central manager prior to matching, or are they transferred directly from the submitter to executor once matched?
Does the same happen for output files, particularly in the case of a Condor-C configuration?
I guess I have to then look at how many jobs could potentially be queued at any one time and the general size of the input/output files.
Does anyone have prior experiences that could help feed into my decision.
HTCondor-users mailing list
To unsubscribe, send a message to htcondor-users-request@xxxxxxxxxxx with a
You can also unsubscribe by visiting
The archives can be found at: