Failover routing is the rule set that carries out payment failover: when the usual route for a transaction is not working, the system sends it down an alternative route without the customer noticing. "Failover" names the general capability; "failover routing" refers to the routing configuration that decides when to switch and where traffic goes.
The logic lives in a gateway or payment orchestration layer connected to more than one provider. It watches for signs of trouble, such as timeouts, error responses or a sudden fall in approvals from one acquirer, and when a threshold is crossed it diverts new transactions to a secondary connection. Once the primary recovers, traffic is moved back, often gradually. The arrangement presupposes acquirer redundancy, meaning that a second provider is already contracted and integrated.
Failover routing responds to outages. It is related to, but narrower than, cascading, which also retries certain declined transactions through another acquirer, and smart routing, which chooses the best route for every transaction on cost and approval grounds even when nothing is broken. A particular risk is duplication: if a request timed out but was in fact approved, resending it elsewhere can charge the customer twice, so systems check the outcome or reverse the first attempt.