Slurm Partitions
The slurm scheduler uses partitions as the method to allocate resources to jobs. The FRCE cluster has several partitions defined, with the major ones summarized below.
| Partition Name | Wall Time Limit | Max Cores Per User | Other Limits |
|---|---|---|---|
| short | 30 minutes | 170 | |
| norm | 5 days | 575 | |
| gpu | unlimited | 350 | 24 P100 GPUs 48 V100 GPUs 4 A100 GPUs 32 L40s GPUs 8 H200 GPUs |
| unlimited | unlimited | 432 | |
| largemem | unlimited | 192 | |
| nci-dragen | unlimited | 64 | 1 Illumina Dragen V3 Server |
Two other partitions will be displayed in the sinfo output. The dragen partition has several servers for the exclusive use by the CCR Sequencing Facility while csbdevel provides a priority channel for a set of programmers who make frequent changes to their code.
A submit script that requests a particular partition might read
#!/bin/bash
#SBATCH --job-name=myjob
#SBATCH --ntasks=8
#SBATCH --time=00:05:00
#SBATCH --partition=short
printf "You are now executing on node %s using {} CPU cores.\n" ${SLURMD_NODENAME} ${SLURM_NPROCS}
Please note the first line, #!/bin/bash in job script is important. Otherwise, SLURM will not let the script be submitted.
To use GPU in the cluster, change the script headers to
#!/bin/bash
#SBATCH --job-name=myjob
#SBATCH --ntasks=8
#SBATCH --time=00:05:00
#SBATCH --partition=gpu
#SBATCH --gres=gpu:p100:1
printf "%s GPU(s) have been allocated with the identification string\n\t%s\n" \
${SLURM_GPUS_ON_NODE} "$(nvidia-smi -L)"