is api gateway the single point of failuretho

hey everyone, ive been thinking about api gateways lately. they seem super useful for managing external access to microservices, handling auth, rate limiting, and so on. but, im concerned about them becoming a single point of failure. if the gateway goes down, everything behind it is inaccessible. how are you guys architecting your systems to mitigate this risk? are there common patterns or tools you use to ensure high availability for your api gateway layer