CISER header image
Cornell University Cornell University CISER

CISER Computing

Making the most of your CISER Computing Account

 

TIPS FOR REDUCING YOUR COMPUTING USAGE COSTS:

  • Log Off, don’t just “disconnect”.  If you are not actively working on a node, or running a job there, be sure to Log off (Start --> Log off). If you close using the “X” in the Terminal Services Taskbar you are only "disconnecting" -- and a number of background processes are left running that are tracked as "application load time" for billing.
  • Close applications (e.g. SAS, Stata, Gauss, Atlas.ti, etc) that you are not currently using . Each open application uses resources that are tracked for billing (including memory & application load time).
  • Stata Users: When using the SET MEM or SET MAXVAR commands, etc, please be sure to request only what you need, and then reduce the memory back to the default level. This may mean you have to reopen your dataset but it will save you a lot in terms of overall memory use. For more detail please refer to our Frequently Asked Questions regarding Stata memory use.
  • Consider joining with others to Purchase a Subscription to CISER Services. Subscriptions may reduce your charges over the pay-as-you-go option. See CISER Billing Frequently Asked Questions for details.

TIPS FOR EFFICIENT USE OF DISK SPACE:

  • Make use of the temporary "scratch" space on each of the CISER Research nodes.  When uploading or downloading large files for temporary use: put them on the T:\ drive (“scratch” space) of any of the RSCH nodes (rsch101 – rsch105).   Then extract only the data you need to save to your home directory.    Delete the temporary files from the T:\ drive when you are done.  Keep in mind that files left on the T:\ drive are not backed up, are not accessible from other nodes, and are deleted after 7 days (or sooner if space is low)
  • Consider joining with others to rent disk space. Our disk is connected to the computing node by fiber channels and the speed of access is so great as to be indistinguishable from the server's local disk. Reading and writing to your own external hard drive or uploading and downloading data from your PC will be much slower and will also use I/O that gets included in your computing usage.