BattaniMetrics, nodes and events
Compare resource usage with capacity and use Kubernetes events to narrow down problems.
A view of cluster health
Overview combines node, Pod, Service and Deployment counts with resource usage. Compare CPU and memory usage against requests, limits and allocatable capacity to identify pressure.
- Ready nodes, running/pending/failed Pods and available Deployments.
- CPU, memory and Pod-capacity indicators.
- Metrics over time, depending on the available data source.
- Node details alongside Kubernetes Events.
Investigate capacity pressure
- Compare usage with total requests and limits in Overview.
- Inspect nodes that are not ready or have high utilization.
- Check namespace events for scheduling, image-pull or volume errors.
- Open the workload to review resource settings and container logs.
Interpret the data source correctly
Live usage depends on access to metrics-server or supported Prometheus data. Historical charts also depend on the available source; short-lived samples do not guarantee long-term retention.
CPU requests, CPU limits and measured CPU usage are different values. If the metrics source is unavailable, do not interpret that as zero actual usage; supplement resource status with events and logs.