DCSC logo
 
ABOUT-DCSC
DCSC/SDU
DCSC/AU
DCSC/AAU
DCSC/DTU
DCSC/KU
 
+Open all         -Close all
 
 
 

 Summary:
 ========

   Each cluster has two queues, 'workq' and 'express'.

   queue 'workq':
   
     grants access to the majority of the nodes in the cluster (all but two nodes).
     There is a cap on the amount of runtime a job can request - it's 200 hours.

   queue 'express':

     queue for testing purposes only - use up to 2 nodes
     as execution nodes. There is a cap of 20 minutes runtime
     per job.

 Details:
 ========

 It's not possible for a user to have more than 250 jobs in the queue.

 The queue destination for a job is specified in one of two ways:

   1) '-q workq' is added on the commandline for the qsub command.

   2) '#PBS -q workq' is added amongst other PBS directives in the 
      pbs job-script (see example).

 To ensure a reasonable and fairly predictable turnaround time for jobs
 submitted to the queues, and make the scheduling strategy more effective
 in managing the resources, the queues have walltime limits on jobs:

   'workq' - jobs may not ask for more than 200 hours

 Walltime limit is specified in one of two ways:

   1) '-l walltime=#Seconds' is added on the commandline for the qsub command.
 
   2) '#PBS -l walltime=#Seconds' is added amongst other PBS directives in the
      pbs job-script (see example).

 An alternative syntax for the time specification is:

   #HH:#MM:#SS  (HH = hours, MM = minutes, SS = seconds)

 ** If no walltime limit is specified the system will set a 10 minutes limit. **