Decommissioning of the HPC cluster
Since January 15th 2025, the historical CNAF HPC cluster has been decommissioned. This section of the guide is kept for some time for reference and will be heavily modified as the HPC bubbles will be installed and made available to users of projects that requested their usage through a formal request to ICSC via the PICA portal.
Down below are listed some frequently used LSF commands and their SLURM equivalents.
Essential commands
| LSF | SLURM | Description |
|---|---|---|
| bsub < script_file | sbatch script_file | Submit a job script to the scheduler |
| bqueues | sinfo | Show available scheduling queues |
| bjobs | squeue -u <username> | List user's pending and running jobs |
| bsub -n 1 -q hpc_int -Is /bin/bash | srun -n 1 -p slurmHPC_int --pty /bin/bash | Request an interactive session |
Job specification
| LSF | SLURM | Description |
|---|---|---|
| #BSUB | #SBATCH | Scheduler directive |
| -q queue_name | --partition=queue_name or -p queue_name | Specify the scheduling queue |
| -W hh:mm:ss | --time= hh:mm:ss or -t hh:mm:ss | Set the maximum runtime limit |
| -We hh:mm:ss | --time-min= hh:mm:ss | Set an estimated runtime |
| -J job_name | --job-name= jobname | Specify a name for the job |
-o filename | --output= filename or -o filename, | Standard job output |
| -e filename | --error=filename or -e filename | Standard job error output |
| -n number-of-cores | --ntasks= number-of-cores or -n number-of-cores | Number of CPU cores |
| -m <host-group-name> | --constraint="<host-group-name>" | To select a node with a specific processor model |
Further commands and options are listed at the following link [32].