kubectl-events

kubectl events [options]

kubectl events is the modern, dedicated subcommand for listing Kubernetes Event objects. It was added as an alpha in kubectl 1.23 and promoted to GA in 1.27, replacing the older kubectl get events form with a cleaner default output and live-watch support.Events are short-lived records emitted by controllers (kubelet, scheduler, controller-manager, custom operators, ...) that describe noteworthy state changes such as pod scheduling, image pulls, liveness-probe failures, or HPA scaling decisions. Use --for to follow a specific Pod, Deployment, or Job, --watch for live debugging, and --sort-by to put the most recent events at the bottom for chronological reading.

-n NAMESPACE

Target namespace (default: current context's namespace).

List events from every namespace.

Restrict to events whose involvedObject matches the given resource (e.g. pod/web-0).

Stream new events instead of exiting after one snapshot.

Comma-separated event types (Normal, Warning).

Sort by a JSONPath field, typically .lastTimestamp or .metadata.creationTimestamp.

Output format (wide, json, yaml, jsonpath, ...).

Suppress the column header line.

Display help information.

Events expire after --event-ttl (default 1 hour); historical incidents need an external aggregator (Loki, Elasticsearch, Datadog) or persistent storage of the Events API. The --for flag is exact: events emitted on a parent resource (Deployment) are not surfaced when watching the child Pod. kubectl events is a separate plugin in some older distributions; on those, fall back to kubectl get events.

kubectl events landed as an alpha command in Kubernetes 1.23 (December 2021), became beta in 1.25, and reached general availability in 1.27 (April 2023). It was contributed to consolidate the various flags previously needed on kubectl get events to make ad-hoc cluster debugging more ergonomic.

kubectl(1), kubectl-describe(1), kubectl-logs(1)