GlideinWMS

Speed up stop/reconf/start
Maybe we wait RH7 or removal of RRD monitor

Originally requested by Factory Operation, consider also the frontend

1. Profile the operations and understand where the time is spent:
- config file operations
- staging area files
- waiting for condor operations
- RRD checks/operations
How does it scale compared to the number of entries/groups?
How much is it affected by age or cruft?

2. Report in this ticket and maybe a document the results of profiling

3. Understand, propose and implement optimizations