Advance notification of downtime, June 2014:
Please be aware that the ICTS HPC clusters (Hex, Hpc and Grid) will be
shut down on the weekend of the 21-22 June 2014 for data center
maintenance. All jobs must be stopped or check pointed by the evening
of Friday 20 June 2014. We hope to have the clusters back up by at
least the morning of Monday the 23 June.
Request for citations:
All users are reminded that as part of the agreement of usage the ICTS
cluster should be acknowledged in any thesis or paper where work
computed on the clusters is referenced, as per: http://hex.uct.ac.za/Citations.html
Space on /home and /scratch:
Users are requested to clean up unwanted files. As space is at a
premium on the cluster we'd like to keep the thresholds for data usage
under 20GB per user on /home/MyUsername. Large data sets should be
moved to /scratch. To find out how much space you are using run the
following command in your /home top level folder:
du . -sh (du space dot space minus sh)
Running of jobs:
Users are also asked not to run software packages directly on the head
node as this will result in excessive load on the head node. All
computational work must be submitted as part of a job via the qsub
command.