[torqueusers] Maui dies

Dave Jackson jacksond at supercluster.org
Tue Oct 26 12:53:14 MDT 2004


Stewart,

  Send us the logs and we will see what we can do.  However, what would
be most useful would be the output of gdb when this failure occurs (see
section 14.6.2.1 in the online Maui documentation at
http://clusterresources.com/mauidocs).

Thanks,
Dave
  
On Tue, 2004-10-26 at 10:11, Stewart.Samuels at aventis.com wrote:
> Folks,
> 
> I have just downloaded the latest snapshot of Maui
> (maui-3.2.6p11-snap.1097510052.tar.gz) and installed in on my system
> running Torque-1.1.0p3.  While running some simple tests, maui dies
> receiving and apparent signal 11 to shutdown.  This has happened to me
> in the past using earlier versions of both Torque and Maui, but I was
> never able to capture the maui log and the problem is not easily
> reproduced.  Additionally, the problem seems to occur when I go into a
> loop to qsub many more jobs than there are nodes or processors to
> process them.  That is, when jobs must wait in the queues for resource
> availability.  If I restart maui, the jobs sitting in the queues get
> scheduled and complete normally.  Oh, something else that might be
> useful in tracking this is that I am running Redhat EL Advanced Server
> 3.0, update 2.
> 
> I know others have seen this problem on earlier versions of maui as
> well.  But has anyone seen it on the latest releases?  Also, the
> maui.log file I have captured is a bit over 2.5 MB in length.  Is
> there someone that can read this log and perhaps help me track down
> the problem?  Any help in trying to track this would be much
> appreciated.
> 
>  
> 
>                Stewart Samuels
>                Technical Advisor
>               Global Unix Engineering Services                
>                ole0.bmp
>            1041 Route 202-206                   
>               Bridgewater, NJ  08807
> 
>               (908) 231-4762
>               Stewart.Samuels at Aventis.com
> 
> 
> 
> ______________________________________________________________________
> _______________________________________________
> torqueusers mailing list
> torqueusers at supercluster.org
> http://supercluster.org/mailman/listinfo/torqueusers



More information about the torqueusers mailing list