Introduction to NCI Gadi: Glossary

SIH facilitates access to compute resources including NCI Gadi

This workshop is a Gadi quickstart

Gadi is a much larger HPC than Artemis

Access is merit-based or paid for

Log in to Gadi via ssh to gadi.nci.org.au

Service units (SU) are the charge rate on Gadi, and projects have a finite allocation of SU/KSU

nci_account -P <project> for KSU per project

lquota for disk and iNode by project

nci-files-report -g <project> -f <filesystem> for disk and iNode by user

Unused KSU are forfeited, and negatively impact future applications

Your Artemis PBS scripts will be easily portable to Gadi, with a few small changes

Gadi has some custom commands and directives

Gadi’s data transfer queue is copyq

Gadi’s data mover node is gadi-dm.nci.org.au

Use Gadi copyq or Artemis dtq depending on your data transfer requirements

Unlike our Artemis support, NCI won’t (rarely) install software globally upon request, but they will assist you to perform local installations

You can make your locally installed software compatible with module commands by following the steps above

NCI monitor job performance. Users with consistently underperforming jobs will be singled out!

Time spent benchmarking and optimising key jobs in your workflow can save you considerable walltime and KSU

As Gadi does not allow job arrays, we recommend OpenMPI and nci-parallel to distribute parallel tasks over CPUs/nodes

Included examples can hopefully be adapted for your own work