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

[condor-users] Preempting Java Universe Jobs Takes Too Long

I've noticed that on Windows machines (runnning Condor 6.5.5 and 6.5.4) I use default configuration on these Windows nodes. The problem is that Condor takes a long time to preempt Java universe jobs. Condor's checkpointing mechanism does not work for Java universe. So, as far as I can understand, preemption for Java jobs means just killing them. Now, if this is true, why does killing a Java job take up to 5 minutes?

Alexander Klyubin

Attachment: pgp00009.pgp
Description: PGP signature