[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[HTCondor-users] Query about working of FLOCK(ing)



Hello Experts,

It's mentioned in Link [1] "A particular job will only flock to another pool when it cannot currently run in the current pool."

I am trying to understand the meaning of above statement:Â

- Submitted a batch 1K jobs all jobs started running randomly on all condor pools configured in condor_host or flock_list. Initially I thought it could be possible that after one Negotiation cycle of pool configured in condor_host it will start considering the pools configured in flock_list but it looks like that sometime both condor_host and pool present in flock_to list simultaneously considering the jobs to run. Also what factors decide the max jobs to be consider for run in single cycle?

What logic is used in background to ensure that it always try to use condor_host first and then flock_to list from left to right? Doc [2] also mentioned about priortizing which is happening at sched level befor negotiator do the match based on priority, does this impact the flocking behavior in any manner?

[1]Âhttps://research.cs.wisc.edu/htcondor/manual/v7.6/5_2Connecting_Condor.html
[2]Âhttps://d1rkab7tlqy5f1.cloudfront.net/EWI/Over%20de%20faculteit/Afdelingen/Software%20Technology/Distributed%20Systems/People/Dick%20Epema/Master%20Theses/MSc-thesis-Evers.pdf


Thanks & Regards,
Vikrant Aggarwal