I'm doing some research on using Akka HTTP + Akka cluster to implement a REST web service that's scalable and resilient to machine failure. The Akka cluster itself is scalable and resilient, but Akka cluster's client (my Akka HTTP application) is not resilient and is a single point of failure (SPOF). Since I'm new to Akka, may I ask what are the common practices to manage High Availability of Akka HTTP? Should I use an external load balancer for that, or am I reinventing the wheel?
Thank you!
Your akka http application shouldn't be a single point of failure. You can run multiple instances and, as you mentioned, run them behind a load balancer.