CLASSE Cluster

From Cornell CAC Documentation

Revision as of 20:29, 8 March 2012 by Jhs43 (Talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

CLASSE Cluster

  • 10 compute nodes [compute-0-[0-9], one head node [classe.cac.cornell.edu] running Red Hat 5.1 linux.
  • Monitor it with Ganglia
  • Runs the maui and Torque resource manager.

Maui Scheduler and Torque Resource Manager (examples from the v4 cluster)

  • Submit HELP requests: help OR by sending email to: help@cac.cornell.edu

Running Jobs

The scheduler has one queue, default. It is set to allow jobs to share nodes and allocate resources per-core. Jobs spread across the nodes first, so the first 10 jobs will run on 10 nodes and, after the next 10 jobs, each node will have 2 jobs running.

Common commands:

  • qsub - submit jobs
  • showq - see running jobs
  • canceljob - cancel jobs

How is My Job?

Given your job ID, try qstat -n or checkjob <jobid> to see what node your job is on. Then go to Ganglia and find that node. Click on the node in Ganglia to get more detailed information.

Retrieved from "http://www.cac.cornell.edu/wiki/index.php?title=CLASSE_Cluster"
Personal tools