Cluster Page

The Cluster page displays a high-level view of the cluster and overall health of workloads and pods. The main panes include:

  • Cluster Pod Map—shows workloads (or deployments) and their pods. When a workload scales up or down, the Cluster Pod Map immediately updates to show the current state.

  • Workloads table—provides more details about each workload, such as its type and health. Health is computed from metrics according to criteria specified in the Monitor.

The Cluster Pod Map can be visually busy when many workloads are running. Use the drop-down in the Controls panel to filter by namespace and simplify the display. This will be more helpful if the workloads of interest are grouped in namespaces.

View all monitors for a workload by clicking it in the Cluster Pod Map or Workloads table. This opens the Workload page that shows monitors and other details, such as the IP address of the node on which the workload is running.

Cluster page

The following provide more details about Cluster page panes and panels. From the top left, moving clockwise, these include:


In the Controls pane, the Grafana icon links to the native Grafana dashboards. Use the Namespace drop-down to filter objects displayed in the Cluster Pod Map and Workloads table.

Cluster Pod Map

The Cluster Pod Map shows the relationship between pods and workloads as well as their current state. The inner circles are pods (hover over one to see its name) and the outer containers (or circles in the case of one pod per workload) are workloads. Workloads and pods are colored to indicate health as computed from their respective monitors. The screen below shows four pods in the es-demo workload, with an overall workload status of green.

Cluster Pod Map

Workloads table

The Workloads table lists the workloads and shows a collection of attributes. The metrics are calculated from the various data exporters used by the workload. Columns include:

  • Service Types: Each workload can have one or more service types. Prometheus scrapes metrics at a regular interval (currently set to 10 seconds) and those metrics are used to determine the set of service types supported by the workload.
  • nodes : pods : containers: column reflects the infrastructure load.
  • Container Starts: Number of containers in this workload that have been started.
  • Health: One of Critical, Warning, Healthy, or Unknown. The health is computed from metrics according to criteria defined in the monitor.

Cluster Health Timeline

The Cluster Health Timeline rolls up the health of all workloads over time.

Cluster Details

The Cluster Details panel summarizes the cluster infrastructure and health and lists the namespaces.