• Clusters! We have …
  • New Configuration

• Tails! We have …
  • greentail (HP) primary login node and scheduler (do not run long running programs on it)
    • current file server, to be backup file server
  • swallowtail (Dell) secondary login node and scheduler (do not run long running programs on it)
    • does some backup
  • petaltail (Dell) a dev/test box (go ahead, crash&burn programs …)
    • does some backup
  • sharptail (Microway) no logins
    • to become file server
  • whitetail
    • access to Hadoop

• Queues! We have …
  • Our Queues

• New file server!
  • sharptail will take over serving /home from greentail
  • greentail will become a backup server (but remain as primary login head node)
  • disk-to-disk backups from array to array will safe guard data
  • greentail will server /sanscratch to all nodes but sharptail will serve /sanscratch to n33-n37
    • 10 TB in total for scratch, use it

• Downtime!
  • one day, between fall and christmas break
  • switch file servers, remount /home
  • some other maintenance activities
  • queues need to idle
  • final sync of data from greentail to sharptail

• Hadoop!
  • Use Hadoop Cluster

• GPU-HPC!
  • Submitting GPU Jobs

• What's Coming!
  • Perhaps replacement hardware for the Dell cluster
  • More powerful servers, much greener, perhaps fewer job slots
  • A physical plant/ITS venture …

Back