Google Professional Cloud Network Engineer Question 734
Select 3Google Cloud PlatformYou are managing a web application that uses an Application Load Balancer (ALB) in Google Cloud. The application is being updated, and you need to safely drain traffic from an old backend service (service A) and redirect it to a new backend service (service B) without causing downtime. Which steps should you take to achieve this?
- A
Set the backend service A's max rate to 0 in the ALB configuration to drain traffic.
- B
Update the ALB's URL map to route all traffic to backend service B.
- C
Remove backend service A from the ALB configuration immediately.
- D
Configure a connection draining timeout for backend service A to allow existing connections to complete.
- E
Disable the health checks for backend service A to stop new requests from being sent to it.
Show answer and explanation
Correct answers: A, B, D
Explanation
To safely drain traffic from an old backend service and redirect it to a new one using an Application Load Balancer, you must ensure a controlled process to avoid service disruption. Setting the max rate to 0 and configuring a connection draining timeout gradually stop traffic to the old service while allowing existing connections to finish. Updating the ALB's URL map ensures new requests are routed to the new backend service. Immediate removal or disabling health checks can cause active connections to fail, leading to downtime.
- A. Correct.
Correct. Setting the max rate to 0 ensures no new requests are sent to backend service A, starting the traffic draining process.
- B. Correct.
Correct. Updating the URL map to route traffic to backend service B ensures new traffic is directed to the new service.
- C. Incorrect.
Incorrect. Removing backend service A immediately would terminate active connections, potentially causing downtime.
- D. Correct.
Correct. Configuring a connection draining timeout allows existing connections on backend service A to gracefully complete before traffic is fully drained.
- E. Incorrect.
Incorrect. Disabling health checks would make the backend service appear unhealthy but does not guarantee a controlled traffic draining process.