squeue & kuacc-queue & valar-queue

After submitting jobs on KUACC or VALAR, users should monitor their jobs to check the status and resource usage.

squeue: Displays information about jobs currently in the Slurm scheduling queue (pending, running, held, etc.).

squeue –u username

Shows only your jobs.

kuacc-queue / valar-queue: Cluster-specific helper scripts that provide a simplified or customized view of the job queue (often with extra formatting or filtering for those systems).

kuacc-queue | grep username

valar-queue | grep username

kuacc-queue | more

valar-queue | more

Below is an example of the valar-queue command output, along with explanations of each column.

	Description
JOBID	Unique identifier assigned to the job by Slurm.
PARTITION	Partition (queue) where the job is submitted, e.g., short, long, gpu.
NAME	Job name (set with #SBATCH -J; if not provided, the script name is used).
USER	Username of the job owner.
TIME_LEFT	Remaining walltime before the job reaches its limit.
TIME_LIMIT	Maximum walltime allocated to the job (from #SBATCH -t).
START_TIME	Actual or scheduled start time of the job.
ST (STATE/STATUS)	Current state of the job. Common values include: R = Running PD = Pending CG = Completing CD = Completed CA = Cancelled F = Failed TO = Timeout NF = Node Fail OOM = Out of Memory S = Suspended
NODES	Number of nodes allocated to the job.
CPUS	Total number of CPU cores allocated to the job.
NODELIST(REASON)	If running, shows the node(s) allocated. If pending, shows the reason (e.g., Resources, Priority, Dependency).

NODELIST(REASON) Column: Meaning of Common Reasons

Reason	Meaning
InvalidQOS	The job’s Quality of Service (QOS) is invalid.
Priority	One or more higher-priority jobs are ahead in the queue; your job will run eventually.
Resources	The job is waiting for required resources (CPU, memory, nodes) to become available.
PartitionNodeLimit	The job’s requested nodes exceed the partition’s current limits, or required nodes are DOWN/DRAINED.
PartitionTimeLimit	The job’s requested runtime exceeds the partition’s time limit.
QOSJobLimit	The maximum number of jobs allowed for this QOS has been reached.
QOSResourceLimit	The QOS has reached a resource allocation limit.
QOSTimeLimit	The QOS has reached its maximum allowed time.
QOSMaxCpuPerUserLimit	Maximum number of CPUs per user for this QOS has been reached; job will run eventually.
QOSGrpMaxJobsLimit	Maximum number of jobs for the QOS group has been reached; job will run eventually.
QOSGrpCpuLimit	All CPUs allocated to the job’s QOS group are in use; job will run eventually.
QOSGrpNodeLimit	All nodes allocated to the job’s QOS group are in use; job will run eventually.

For a complete list of reason codes, refer to this link. In practice, the most commonly encountered reasons are Priority, Resources, and QOSMaxCpuPerUserLimit.