Hi All. I am running GE 8.1.2 and I have a situation where once in a while ( 2x a week ), Grid Engine forgets about one of the Subordinate queues.
Everything works as expected where my subordinate queue goes to "S" suspend-mode when a job enters the queue it is subordinate to. However once in a while Grid Engine Forgets and does not suspend and the node is overloaded.
The subordinate setup is correct as this only happens once in a while: # qconf -sq tw | fgrep subordinate subordinate_list free64=1 My fix has been to set it back to "none": # qconf -sq tw | fgrep subordinate subordinate_list NONE then back to how it was: # qconf -sq tw | fgrep subordinate subordinate_list free64=1 And all is back to normal and the job which was supposed to have been suspended which was not suspended is now suspended correctly and the node is not overloaded. Other than running the debugger, anyone else seen this and/or any hints of what may be wrong in my setup? _______________________________________________ users mailing list [email protected] https://gridengine.org/mailman/listinfo/users
