HTCondor Chron Implementation
PRIMARILY: hooks for end of injection jobs
- Add ids to something, have a chron job watch all of them... and run catalog and testgr based on this instead of having it run every time an offsource run finishes.
General Ops:
- Request next (*2) increment of memory or disk if HOLD message says we ran out
- If a folder doesn't exist, try releasing job first; then write it
- see if there's a way to catch a job before it evicts; given fixed number of use-cases
Edited by John Sullivan