Queue Explanation

From UMaine Supercomputer

Jump to: navigation, search

For quick reference, here is a list of the queues on the system and their parameters. All these queues can be accessed directly from the new gateway, panopticon, as well as the compile nodes (fawlty, bender and kearney2). When using the reboot queues be sure to add at least 20 minutes to your wall clock request.

  • linux-spool/darwin-spool
    • Boots into either Linux or Darwin upon request.
    • Access to the Xserve G5's
    • Priority of 1.
    • Maximum of 1 job per user.
    • Maximum Wall Clock time of 48 hours.
  • linux-batch/darwin-batch
    • Boots into either Linux or Darwin upon request.
    • Access to the Xserve G5's
    • Priority of 100.
    • Maximum of 92 nodes (192 processors) per research group.
  • kearney
    • Access to the old Kearney PIII nodes.
    • No restrictions currently.

PBS Script Commands

To select one of these queues, you will need one of the following lines in your PBS scripts:

#PBS -q linux-spool
#PBS -q darwin-spool
#PBS -q linux-batch
#PBS -q darwin-batch
#PBS -q kearney
Personal tools