Skip to content

Queue Status

The Queue Status page provides a consolidated view of compute queue availability and job statuses within a selected cluster. It allows users to monitor node, CPU, vCPU, and memory resources along with currently running and pending jobs.

Queue Overview

Accessing the Queue Status

To access the Queue Status page:

  1. Navigate to Queue Status from the left menu.
  2. Use the filter options at the top to select:
  3. Provider: Select the cloud provider, such as AWS.
  4. Account: Select the cloud account.
  5. Region: Select the region.
  6. Cluster: Choose the cluster you want to monitor.
  7. Duration: Select the date range.
  8. Click Search to load the relevant queue data.

Queue Filter

Overview Panel

This section displays doughnut charts representing cluster-wide resource limits:

  • Max Nodes — Yellow
  • Max CPU Cores — Blue
  • Max vCPUs — Purple
  • Max Memory (GB) — Orange

Queue panel

  • By default, the charts display the maximum capacity, shown as Max values.
  • Hover over a chart to view its tooltip, including Available values such as available nodes or cores.

Available Value

Queue Panel (Left Sidebar)

All available queues within the cluster are listed with resource availability indicators. Each queue displays:

  • Queue name, such as gpu, large, medium, or mlgpu.
  • Cores Available.
  • Usage Percentage.

Queue panel

Running Jobs

The Running Jobs panel shows a list of currently active jobs.

Job Details Displayed

For each running job, the following details are shown:

  • Submit Time (UTC): The timestamp when the job was submitted.
  • Job Name: The name assigned to the job.
  • Job ID: The unique identifier of the job.
  • Submitted By: The user who submitted the job.
  • Requested Resources:
  • Number of CPU cores.
  • Memory in GB.
  • Node count.
  • Estimated Wall Time: The maximum time the job is allowed to run.

Running Job Status

Pending Jobs

The Pending Jobs panel shows jobs that are queued and waiting for execution in the selected queue.

Job Details Displayed

For each pending job, the following details are shown:

  • Submit Time (UTC): The timestamp when the job was submitted.
  • Job Name: The name assigned to the job.
  • Job ID: The unique identifier of the job.
  • Submitted By: The user who submitted the job.
  • Requested Resources:
  • Number of CPU cores.
  • Memory in GB.
  • Node count.
  • Estimated Wall Time: The maximum duration for which the job is allowed to run.
Info:
  • Click Search to refresh the displayed values.
  • Hover over the doughnut charts to view resource availability details.
  • Use this page to review queue congestion and resource availability.
General Notes:
  • If you encounter UI issues or have questions about job submissions or status, contact your administrator for support.
  • Ensure you have selected the correct provider, account, region, and cluster when filtering queue data.