Platform

Get Cluster Event

Changed on

Required role: read or higher.

Get specific cluster monitor event.

The identifiers of the events can be retrieved via GET /v0/cluster/events. At most 1000 events are retained (newest first), and events older than 72h are deleted. The latest event, if it already exists, is never cleaned up. This endpoint can return a 404 for an event that no longer exists due to clean-up.

get/v0/cluster/events/{event_id}

Request

  • The document declares no server URL.
  • Auth: HTTP bearer

Path parameters

event_idstring required

Cluster monitor event identifier or latest

Query parameters

selector'all' | 'status'

The selector parameter limits which fields are returned. Limiting which fields is particularly handy for instance when frequently monitoring over low bandwidth connections while being only interested in status.

Response

all_healthyboolean required

Whether every service was healthy when this event was recorded. Says nothing about the cluster now: for that, read stale alongside it or call GET /v0/cluster_healthz.

api_resources_infostring nullable
api_self_infostring nullable
api_status'InitialUnhealthy' | 'Unhealthy' | 'Healthy' required
compiler_resources_infostring nullable
compiler_self_infostring nullable
compiler_status'InitialUnhealthy' | 'Unhealthy' | 'Healthy' required
idstring uuid required

Cluster monitor event identifier.

recorded_atstring date-time required
runner_resources_infostring nullable
runner_self_infostring nullable
runner_status'InitialUnhealthy' | 'Unhealthy' | 'Healthy' required
staleboolean nullable

Whether the monitor stopped writing, making the statuses below the last recorded ones. Reported for the latest event only: older events are old by design.

Changes