Skip to content

Accessing Kestrel RHEL9#

RHEL9 Login Nodes#

To submit jobs to RHEL9 compute nodes, you must be connected to an upgraded login node. Use the following login nodes to access RHEL9 resources:

  • CPU login node: kl3.hpc.nlr.gov
  • GPU login node: kl5.hpc.nlr.gov
  • DAV node: kd7.hpc.nlr.gov

Example login#

The following example describes how to access the RHEL9 CPU login node.

1. SSH to the appropriate login node.

ssh kl3.hpc.nlr.gov

cat /etc/redhat-release
lscpu | grep 'Model name'

Example output:

Red Hat Enterprise Linux release 9.4 (Plow)
Model name: Intel(R) Xeon(R) Platinum 8470

2. Verify the currently loaded default modules.

module list

Example output:

Currently Loaded Modules:
   1) Core/26.05   2) application/26.05   3) DefApps

3. Verify the installed Slurm version.

srun --version

slurm 25.05.6

4. Launch an interactive job. A compute node running RHEL 9 is assigned automatically.

salloc --time=20:00 --account=<allocation handle> --nodes=1

Example output:

salloc: Granted job allocation 13338420
salloc: Waiting for resource configuration
salloc: Nodes x1005c0s0b0n0 are ready for job

Then verify OS on the compute node:

cat /etc/redhat-release