Question

After deploying a new version of a web application behind a Layer 7 load balancer, all backend instances are marked unhealthy and the load balancer returns HTTP 502 to users. The instances are running and respond with HTTP 200 when tested directly via curl on their private IP at '/'. The load balancer health check is configured to probe '/healthz' on port 8080. A cloud engineer confirms the application listens on port 8080 but the new release moved the health endpoint to '/health'. Which action will most directly restore healthy status?

Accepted Answer

Update the load balancer health check path from '/healthz' to '/health' to match the application's new endpoint

Answer

Increase the health check timeout and unhealthy threshold so instances have more time to respond

Answer

Restart all backend instances to force them to re-register with the load balancer

Answer

Add an inbound security group rule allowing port 8080 from the load balancer subnet