kuacc-info

The

CODE

kuacc-info

command provides a comprehensive summary of resource usage on the KUACC Cluster, including CPU, memory, and node utilization.

The output consists of two main sections:

1. User Summary

The first section lists all users with their currently running and pending jobs.

It also shows how many nodes, CPUs, and how much memory (in GB) each user is consuming across different partitions:

COSMOS – Nodes dedicated to COSMOS users or workloads
ILAC – Nodes dedicated to ILAC users or workloads
AI – Legacy GPU nodes for AI workloads

Example (excerpt):

CODE

USER RUNNING PENDING COSMOS|ILAC|AI COSMOS|ILAC|AI COSMOS|ILAC|AI 
aghalhar22 1 0 0| 1| 0 0| 1| 0 0.0| 64.0| 0.0 
eyasar25 2 0 1| 0| 0 2| 0| 0 8.0| 0.0| 0.0 
ykhojakulov20 5 0 0| 2| 0 0| 5| 0 0.0| 80.0| 0.0

2. Node Details

The second section lists all available nodes in the cluster, showing detailed information about their CPU and memory utilization, partition, and status. This section serves as an extended version of the

CODE

sinfo

command, providing a more detailed per-node view.

Example (excerpt):

CODE

------------------------------------------------------------------------------------------------------------ 
NAME TYPE CPU_USAGE MEM_USAGE STATUS 
------------------------------------------------------------------------------------------------------------ 
be01 ILAC 2 ( 16%) 32.00 ( 26%) BUSY 
be02 ILAC 0 ( 0%) 0.00 ( 0%) FREE 
da02 BIYOFIZ 16 ( 80%) 64.00 ( 26%) BUSY 
dy02 AI 8 ( 22%) 256.00 ( 52%) BUSY 
ke03 COSMOS 0 ( 0%) 0.00 ( 0%) FREE

Usage

The

JAVA

kuacc-info

command is particularly useful for users planning to submit new jobs.

By inspecting the STATUS column, users can identify available (FREE) nodes and use them for job submissions via the

JAVA

--nodelist

or

JAVA

--constraint

flags in SLURM.

CODE

sbatch --nodelist=ke03 job_script.sh