I have an experiment with nested DAGs and thousands of submit files.
Condor seems to associate one submit file to one job, including
condor_dagman invocations.
Is there an easy way I can see the nested hierarchy of jobs in my DAG
tree? Or somehow get a higher-level grouping of all the jobs
associated
with one of my "experiments"? I'd like this functionality so as to
run
multiple experiments at once, and to kill off experiments that have
problems/bugs that are discovered after invocation.
I could probably do this by parsing the log file for the highest-level
DAG submit file, but I'd have to be a little clumsy about making sure
the experiment is still running, and cross-reference that list with
what
jobs are actually running. I'm just wondering if there's a simpler
way
to do this.
Essentially I'm looking for a 'pstree'-equivalent for Condor (where
condor_dagman invokes nested jobs). Has anyone written such a utility
and is willing to share?
Thanks,
- Armen