AZ104-Q: stop sending traffic to a failed VM

A TCP service runs on two VMs behind a Standard Load Balancer. One VM’s application fails but the VM stays running. What should determine whether it receives new traffic?

  • A) A health probe that checks the application endpoint. ✓
  • B) A resource lock.
  • C) A DNS CNAME record.
  • D) A storage lifecycle rule. Answer: A. A health probe can remove an unhealthy backend from load-balanced traffic. Why not the others: Locks, DNS aliases, and lifecycle rules do not assess backend health.

load-balancer · health-probe