Skip to main content
Version: FCP 25.11

Node Monitoring

The node monitoring panel provides detailed status and resource usage for all compute nodes in the cluster, and also displays a cluster-level job list. With this panel, you can view node health, resource consumption, and job execution status in near real time.

Prerequisites​

  • You have a Desktop Portal account and are logged in.
  • You have at least one authorized cluster. Contact your administrator if needed.

Node List​

The node list shows key information for all nodes in the cluster, helping you quickly understand basic node status. The list includes:

  • Node name: unique identifier. Click to open the node's detailed monitoring page.
  • Node ID: unique node identifier in the cluster.
  • Cluster ID: unique identifier of the cluster the node belongs to.
  • Partition: compute partition the node belongs to.
  • Uptime: time since the node last started.
  • CPU count: total CPU cores on the node.
  • Total memory: total memory capacity.
  • Root filesystem: total storage capacity of the root filesystem (/).
  • CPU usage: average CPU usage on the node.
  • Memory usage: percentage of memory used.
  • Root filesystem usage: percentage of used space on /.
  • Swap usage: swap utilization.
  • Scheduler node state: node state in the scheduler, such as idle, allocated, down, and so on.
  • Session count: number of active sessions on the node (SSH/VNC/etc.).
  • User count: number of distinct users with active sessions on the node.
  • Running jobs: number of jobs (tasks) currently running on the node.
  • Scheduler CPU total: total CPU cores configured in the scheduler for this node (may differ from physical cores due to oversubscription policies).
  • Idle CPUs: number of free CPU cores not used by jobs.
  • Running CPUs: number of CPU cores currently used by running tasks.

View Jobs​

This section shows job status across the entire cluster and provides quick filter tabs to help you view jobs in different states. For more information, see Job List.