[HTCondor-users] Determining the full path of the initial dir of the DAG?

[Michael Pelletier <Michael.V.Pelletier@xxxxxxxxxxxx>]

I'm working out the final details of a DAG, and need to supply each job node with a path to its input file at the top level of the DAG directory, even though each job node goes one deeper into a directory tree built by each job as the DAG works through the branches of an optimization problem.

Is there a way to get the full path of the directory from which condor_submit_dag was run? I'd like to add it to VARS to pass into the job, or figure out if there's something already being populated. I haven't been able to find anything in the manual thus far. I think I can kludge something applicable to all nodes by doing something along these lines in the job submission relative to the DAG node's deep initial directory as it works its way down the branches via an include command:

ls ../input2.dat ../../input2.dat ../../../input2.dat ../../../../input2.dat

But if I could just give the arguments a full path to the DAGroot/input2.dat file, that'd be far more sensible-looking and self-documenting.

Thanks!

Michael V. Pelletier
Information Technology
Digital Transformation & Innovation
Integrated Defense Systems
Raytheon Company