Skip to content
HomeUsageHealth Diagnostics

Health Diagnostics ​

Health Diagnostics runs a set of built-in checks against a cluster and turns the results into a single health score with actionable findings. It answers the question "is this cluster healthy right now?" without writing your own queries.

Health Diagnostics

What the report contains ​

  • Health score — an overall 0–100 score derived from the check results.
  • Findings — grouped into four categories, each with a severity (critical, warning, info), an occurrence count, and a sample:
    • nodes — NotReady nodes, resource pressure, and similar node-level issues.
    • workloads — failing or restarted pods, unavailable replicas.
    • storage — PVC binding and capacity problems.
    • platform — control plane and add-on health.
  • Control plane components — ready/total counts for API server, scheduler, controller manager, and etcd.
  • Top consumers — nodes and pods with the highest CPU and memory usage. This section is only shown when Prometheus is enabled for the cluster.

Multi-cluster overview ​

The overview page aggregates the health of all clusters you can access, so you can spot the unhealthy cluster in a fleet at a glance.

API ​

MethodPathPurpose
GET/api/v1/health-reportHealth report for one cluster
GET/api/v1/clusters/overviewAggregated multi-cluster overview

TIP

Health checks are computed on request, not stored — every page load reflects the current state of the cluster.

Released under the Apache License.