Running Jobs on Māui and Mahuika
The section describes the key features of the Slurm multi-cluster configuration on Māui and Mahuika, specifies the queues (or partitions) to which you can submit jobs, provides tips on how to optimise your job throughput, and provides a cheat sheet for some oft-used Slurm commands.
- GPU use on NeSI
- Job Checkpointing
- Slurm Interactive Sessions
- Fair Share
- Job prioritisation
- Mahuika Slurm Partitions
- Milan Compute Nodes
- Māui Slurm Partitions
- SLURM: Best Practice
- Hyperthreading
- NetCDF / HDF5 file locking
- Checksums