# Multi-target Failover (https://portunus.bybee.dev/en/docs/overview/concepts/multi-target-failover)



> **How to set it up:** [Standalone (TOML)](/en/docs/standalone/forwarding-rules#multi-target-failover) · [Server + Client (operator)](/en/docs/server-client/forwarding-rules#multi-target-failover)

Available since v0.7.0. Extends `Rule.targets[]` from a single target to
an ordered list of up to 8.

## How it works [#how-it-works]

* **Lower priority number = higher preference**. The client always
  attempts the highest-preference healthy target first, then walks the
  remaining targets in priority order on failure.
* **Passive failure detection**: 3 consecutive connect failures within
  30 s flips a target to `Failed`; 2 consecutive successes flip it back.
* **Active probes** (optional, per rule): set
  `health_check_interval_secs` (range `1..=3600`) to drive periodic
  TCP-connect probes from the client.
* **All-failed fallback**: when every target is `Failed`, the client
  still attempts the highest-priority target rather than giving up.
* All health state is **client-side and ephemeral** — it does not
  survive a client restart.