CockroachDBPrometheus

CockroachDB Range Unavailability Before Quorum Loss

critical
reliabilityUpdated Feb 12, 2026

CockroachDB ranges with fewer live replicas than needed for quorum (cockroachdb.ranges_replication_problem with unavailable ranges) indicate impending data unavailability. This is the critical pre-failure signal before queries start failing due to lost quorum.

How to detect:

Alert immediately when cockroachdb.ranges_replication_problem shows unavailable ranges > 0. This metric indicates ranges that cannot serve reads or writes due to insufficient live replicas for quorum.

Recommended action:

Investigate node health, restore failed nodes immediately, or adjust replication factor if consistently under-provisioned. This requires urgent intervention to prevent data unavailability.