Configure load balancers

Our load balancer config needs an update before Monday. Apply the new backend pool that adds web03 and web04, set health checks to expect HTTP 200 on /healthz with a 3s timeout, and make sure session affinity is cookie-based. After you update, validate routes from the edge and roll back if latency increases by more than 30ms for the VIP.

Before I give this to engineering for approval, make the proposed load balancer change easy to accept: put the expected traffic distribution and capacity headroom up front, list the exact URL and health-check response used, show how session affinity affects sticky sessions for the mobile app, and call out what a reviewer must test in staging to be confident.

I added web04 to the pool and it immediately failed health checks. I’m worried traffic won’t shift properly and mobile users will see session loss. I don’t know whether the service on web04 is misconfigured, the health endpoint is wrong, or networking is blocked. What diagnosis path will quickly tell me the cause and the safest way to bring web04 into rotation?

Every few months we scramble because new backends fail health checks or affinity breaks, forcing rollbacks and late-night fixes. Where am I losing time and credibility in how we add servers, and what single habit change would prevent most of these emergency rollbacks?