[torqueusers] Torque on 1000 nodes ?

Dave Jackson jacksond at clusterresources.com
Fri Jul 1 16:06:00 MDT 2005


Ole,

  Garrick's comments are correct.  However, in his note, he described a
number of changes made for his local site to support local needs.  Some
of these were unrelated to increasing support for job count.  Some were
to improve support for a large number of nodes, others were for support
of a large number of classes.

  In the latest releases of Maui, only the MMAX_JOB #define must be
modified to increase max job support.  In earlier releases which have
the integrated 'MCom' library, both MMAX_JOB and MAX_MJOB must be set.
The Maui docs have been updated to reflect this.

Thanks,
Dave



On Fri, 2005-07-01 at 23:07 +0200, Ole Holm Nielsen wrote:
> Hi Dave,
> 
> Nice to hear from you again.  I did read about increasing the hardcoded
> limits in Maui,
> http://clusterresources.com/products/maui/docs/a.ddevelopment.shtml
> However, it seems that several additional parameters need to be edited
> according to Garrick Staples, see
> http://www.supercluster.org/pipermail/torqueusers/2005-June/001638.html
> 
> Would you kindly check if Garrick's point is correct and the Maui docs
> Appendix D needs correction ?
> 
> Thanks,
> Ole
> 
> >   Maui's maximum job count is hard-coded/static but it is configured as
> > a single #define in a header file and the documentation describes where
> > and how to change this parameter.  There are sites out there running
> > Maui with ~15,000 jobs queued/running and Moab is being evaluated at 3x
> > that.  The bottleneck is usually not Maui/Moab, its typically the
> > resource manager.
> 



More information about the torqueusers mailing list