Daily review flow
- Start from Cluster Overview to identify the object, namespace or event that stands out.
- Open the related drawer to confirm current status, ownership, recent activity and connected objects.
- Use the focused guide below for the object type you are reviewing.
- Move into Troubleshooting when the object is degraded or warning volume is rising.
Operator actionDo not stop at a single status badge. Review related events, owner workload and namespace scope before deciding whether the object is truly unhealthy.
Focused workflow pages
How to inspect a pod
Read readiness, restarts, image, node placement and the next kubectl-style follow-up.
How to read workload readiness
Interpret desired versus ready, rollout health, owner relationships and exposure.
How to review node health
Inspect pressure conditions, allocatable values, runtime data and related scheduling clues.
How to interpret warning events
Judge event severity, timing, repetition and whether the object is actually deteriorating.
How to trace related objects
Follow the chain from pod to owner, node, namespace, events and image metadata.
How to review image metadata
Read tags, digests, immutability, update windows and image consumers.
When to switch to troubleshooting
| Signal | Meaning | Next page |
|---|---|---|
| Restart growth | Pod is changing state, not just carrying old warnings. | Pod and workload triage |
| Desired and ready diverge | Workload may be rolling, blocked or partially unavailable. | Workload readiness |
| Pressure conditions | Node-level saturation may explain scheduling or eviction symptoms. | Node health |
| Repeated access or refusal events | Policy or routing issues may need security or access review. | WS110 / WS111 / WS112 access behavior |