I currently have my own personal pool set up with several workers. A collaborator has offered to add a few powerful workers under the condition that his jobs would get highest priority (can kick my jobs running on his workers) but I can use his workers when they are idle. Does anyone have any advice on how to implement this?
I was thinking of setting up two head nodes (nodes that have
COLLECTOR, MASTER, NEGOTIATOR, SCHEDD)
then and use the flocking mechanism to allow jobs to transfer between the two pools (my pool and my colleague's pool?) The two head nodes would have to be on different machines right? condor_submit has a pool option that must be used with a -name, how do i find out the name of the pool? since im specifying both the sched location using -name and also the pool name using -pool, is there a case where a sched has multiple pools?