View events for a Kubernetes resource
Open the resource inspector and choose Events to see what Kubernetes recorded about that object: scheduling, image pulls, probe failures, restarts and so on. DevOpsPad shows events whose involvedObject.uid equals the object’s UID, newest first.
Matching on UID instead of name matters when a Pod is recreated with the same name. You see the events of the object you opened, not those of an earlier one.
Read the table
| Column | Meaning |
|---|---|
| last seen | lastTimestamp, or eventTime, or the event’s creation time, whichever exists |
| type | Normal or Warning |
| reason | Short code such as BackOff, FailedScheduling or Pulled |
| message | The full event message |
| count | How many times the event repeated |
Click Refresh events in the footer to load them again.
You can also list events directly
get events in a query tab lists every event in the namespace, not only one object’s. Use it with Filter results… to search across a namespace, for example for Warning or a node name. See List Kubernetes resources.
Limits
- At most 200 events, and at most 1 MiB of event data, per object. When either limit is reached, the tab says “Event limit reached; this list is incomplete.”
- Events expire according to the cluster’s retention settings. “No retained events for this resource.” usually means they have already expired.
- The list is a snapshot. It does not update live.
- An object without a UID cannot be matched safely, so no events are shown for it.
Permissions
The Events tab needs list on events in the object’s namespace. If that is missing, the tab shows an access-denied message and the other tabs keep working. See Kubernetes permissions DevOpsPad needs.
Frequently asked questions
Why are there no events for my Pod?
Kubernetes deletes events after a retention period, often about an hour by default. An empty list usually means nothing was recorded recently, not that DevOpsPad failed to read them.