Configure Redis Sentinel to automatically fail over a monitored primary
domain: redis.io · 5 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
Deploy at least 3 Sentinel instances on independent hosts so failure detection has a real quorum rather than a single point of failure
In each sentinel.conf, add sentinel monitor <name> <primary-ip> <primary-port> <quorum>, where quorum is the number of Sentinels that must agree the primary is down
Set sentinel down-after-milliseconds <name> <ms> to control how long a primary must be unresponsive before a Sentinel marks it subjectively down (SDOWN)
Set sentinel failover-timeout <name> <ms> and sentinel parallel-syncs <name> <n> to control failover pacing and how many replicas resync in parallel with the new primary
Point clients at the Sentinel set (not a static primary address) so they can query SENTINEL get-master-addr-by-name and follow failovers automatically
Known gotchas
A primary is only marked objectively down (ODOWN) and eligible for failover once enough Sentinels independently report SDOWN — a quorum of 1 or a misconfigured quorum defeats the purpose
Sentinel handles primary/replica-set failover, not Redis Cluster hash-slot resharding — it's a different mechanism from Redis Cluster's own CLUSTER FAILOVER
Clients that cache a resolved primary address instead of re-querying Sentinel after a failover keep writing to the demoted, now read-only node
Give your agent this knowledge — and 15,500+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?