I try to find out, if it is possible to configure an AWS Application Load Balancer in a way, that if the backend service responds with an error status code (e.G.
503 Service Unavailable) to a regular incoming http request, that the Application Load Balancer then marks the requested backend service as "unhealthy"? Maybe even tries a different one?
I am aware of the typical health checks that AWS Application Load Balancers do. But they only occur in specific intervals. Meaning that if in between of these intervals something goes bad, all upcoming requests until the next interval would still be routed to the already broken backend service.
I am looking for a similar behavior as it is describe for nginx and apache in this post: Load balancer to handle server errors silently