[gridengine users] master node selection and $fill_up behaviour revisited

Dave Love d.love at liverpool.ac.uk
Tue Jun 28 16:23:23 UTC 2011


Michael Weiser <M.Weiser at science-computing.de> writes:

> Hello,
>
> in July 2010 I asked on the users mailing list back at SunSource about a
> peculiar regression in master node selection behaviour of SGE 6.2u5.
> (see http://markmail.org/message/svuskq5qc6oe3axv) After some discussion
> Andy pointed out that I was most likely hitting IZ 3148 which was fixed
> in 6.2u6. And indeed, I was not able to trigger the bug in 6.2u6, which
> was worst of all, because I couldn't upgrade.

62patches.txt says

  3148 6882584 parallel jobs do not always go to the least loaded host 

and the Univa repo has something which looks the same, but we don't have
access to the actual issue to know for sure:

  commit f9022320f5c48fb6de9bba756264e8395e174527
  Author: Joachim Gabler <JGabler at univa.com>
  Date:   Mon May 16 09:57:58 2011 +0200
  
      JG-2011-05-16-0: Bugfix:      parallel jobs are not dispatched to the least
                                    loaded host
                       Issue:       GE-3479
                       Review:      EB
  
> Today I've tried a recent build of V800_BRANCH of
> https://github.com/gridengine/gridengine.git and was able to reproduce
> the bug just as with SGE 6.2u5.
>
> Does anyone here have a handle on the issue and can help out in tracking
> it down and fixing it?

(I'm not sure I understand the circumstances around the issue.)

> Does perhaps one of the other forks fix the bug?

No, if that change doesn't.

-- 
Excuse the typping -- I have a broken wrist



More information about the users mailing list