pv_testMgr is dependent on mdiag

[cadejager]

pv_testMgr is a script we use to keep a cluster full of jobs while testing. The current implementation checks to see how many jobs are currently runnning using mdiag. This is necessary so that it can know if it needs to submit more jobs. For our current state we need to support slurm.

I have already begun work on this in commit a6c896e. This may be all that is needed but I need to test on a non production system.

Pavilion has some work within it to do this which would allow the testMgr to work with any scheduler hopefully but there is just a comment that says it is not working.