[torqueusers] Re: qsub problem in 2.3.2?

Tom Rudwick tomr at intrinsity.com
Fri Aug 15 18:27:32 MDT 2008


This was apparently a side-effect of our upgrading from an older version (2.3.1)
and migrating existing jobs to the new version. When testing the same rpms
on a test network without queued/running jobs, we could not reproduce this
problem.

So if you have the luxury of being able to empty out all queues, it is probably
a good idea to do so.

Tom

Tom Rudwick wrote:
> Is it just us, or did this break in 2.3.2?
> 
> qsub -l mem=5gb test
> 
> When we do this with 2.3.1, we get a job with 5gb memory allocation.
> 
> When we do this with 2.3.2, we get a job with the server default memory 
> allocation.
> 
> Thanks,
> Tom
> 



More information about the torqueusers mailing list