The Rāpoi HPC Cluster (aka raapoi) is a Uni-wide computing resource that uses the Slurm resource manager to schedule jobs and reserve resources. Similar to most modern compute clusters, raapoi requires you to request CPU, Memory and Time for your job. If you do not request these resources, you will be given the minimal defaults (1 CPU, 1 GB Memory, 1 hour runtime), which may not be enough to run your job. The good news about resource reservations is that the resources you request are guaranteed to be yours, the bad news is if you request too little memory or time, your job may terminate prematurely and if you request too few CPUs then your job may run slowly.

Raapoi is made up of partitions. A partition is a set of compute nodes (servers) and each partition has its own configuration and hardware profile. The partition on which you run your jobs will depend on the type of workflow or job you intend to submit.

The cluster employs the module environment to allow researchers to customize their environment with their required applications and languages.

The documentation contained in this wiki cover most of what you will need to know to start running jobs, but if you need more help, we will be creating some tutorials with step-by-step instructions for our most popular apps and languages.