[GE issues] [Issue 3216] rerun of a tightly integrated parallel array job crashes qmaster after restart

joga Joachim.Gabler at sun.com
Wed Dec 23 09:36:08 GMT 2009


------- Additional comments from joga at sunsource.net Wed Dec 23 01:36:06 -0800 2009 -------
This is a duplicate of IZ 1416 - still I'll keep reporting into this IZ,
as we have much more information in here, and might want to fix it in a different way.

The bug was reintroduced in 6.2u2.

The fix for IZ 1416 was to disable reducing of pe task data (the quick fix proposed above).
A proper fix would have to make sure that the event master does proper filtering
when delivering the sgeE_JOB_LIST event (total update on the job list), 
which apparently never worked.


To unsubscribe from this discussion, e-mail: [issues-unsubscribe at gridengine.sunsource.net].

More information about the gridengine-users mailing list