[torqueusers] torque/maui hanging bug(?)

Will Nolan will at headlandstech.com
Mon Aug 9 16:38:17 MDT 2010


>I would say you got the right list.  How often do you get hung? Did you 
>patch your copy of the code?
>
>If you have patched your code and it works for you I suggest submitting 
>a bug to www.clusterresources.com/bugzilla and posting the patch there.
>
>Ken Nielson
>Adaptive Computing

Maui got hung up pretty consistently.  I was able to reproduce it every time with my test case of dumping a large amount of jobs onto torque all at once.

I patched my copy of the code locally and ran it with some debugging printouts that, in conjunction with the maui logs, let me verify that it actually worked (apart from the stand-alone test program I wrote to test my initial hypothesis).  I have been hammering our local torque install with jobs ever since, without any issues.

I just went and cleaned up the code to remove the debugging info and comments, re-built, and re-tested -- all checks out.  I will post the bug to bugzilla along with my patch.  Hopefully it can get field tested a bit more...

Thanks,
Will



More information about the torqueusers mailing list