Skip to content

Observer health

Observers — projections, reducers, and reactors — consume event sequences to build read models and trigger side effects. These tools let an agent check their health, find failures, and surface maintenance the store recommends.

ToolDescription
list_observersList observers in a namespace, with health and replay status. Optionally filter by type (reactor, reducer, projection, or all).
get_observerShow detailed information about a specific observer — its type, running state, owner, handled sequence numbers, and observed event types.
list_failed_partitionsList observer partitions that have failed and are paused.
list_recommendationsList active maintenance recommendations the store has raised.

When an observer is stuck or a read model looks stale:

  1. list_observers — check the observer’s running state and how far its sequence numbers have progressed.
  2. get_observer — inspect the observer in detail, including which event types it observes.
  3. list_failed_partitions — see whether a partition has failed and paused the observer.
  4. list_recommendations — see whether the store already recommends a fix.

A quarantined observer does not auto-resume; an operator has to clear the quarantine. Use the Cratis CLI or the management API to replay, retry, or clear quarantine once you have diagnosed the cause here.

get_observer shows the event types a single observer consumes. For the reverse — every consumer of a given event type across the store, plus its fields — use the design-time event catalog.