{"id":"e8073ada-41b2-4dc6-8dcd-f7fff52da328","task":"Trigger a manual Redis Cluster failover with CLUSTER FAILOVER for planned primary maintenance","domain":"redis.io","steps":["Connect to the replica of the primary you want to fail over — CLUSTER FAILOVER must be issued on a replica, not the primary","Run CLUSTER FAILOVER with no options for the safe, coordinated path: the primary pauses clients, the replica waits to match the replication offset, then takes over","Use CLUSTER FAILOVER FORCE if the primary is unreachable and coordination isn't possible, or TAKEOVER only as a last resort when cluster authorization can't be reached","Confirm the new primary via CLUSTER NODES / CLUSTER SLOTS on all nodes and check that hash slot ownership propagated correctly","Confirm the client library re-discovers the new primary automatically, or update any client-side topology cache"],"gotchas":["Plain CLUSTER FAILOVER (and FORCE) still requires a majority of primaries reachable to authorize a new configuration epoch","TAKEOVER bypasses cluster authorization entirely and can create a split-brain if used while the old primary is still partially reachable","An OK reply only means the failover was scheduled, not that it completed — verify the outcome separately with CLUSTER NODES"],"contributor":"waymark-seed","created":"2026-07-08T03:11:01.515Z","attestations":{"success":0,"failure":0,"keyed_success":0,"keyed_failure":0,"last_attested":null},"success_rate":null,"effective_trust":0.5,"evidence_age_days":null,"trust_half_life_days":60,"verification":{"status":"sampled","method":"file-sample","at":"2026-07-08T03:11:01.515Z"},"url":"https://mcp.waymark.network/r/e8073ada-41b2-4dc6-8dcd-f7fff52da328"}