Google Professional Cloud Developer Question 363
Select 3Google Cloud PlatformYou are deploying a stateless microservice application to Google Kubernetes Engine (GKE). The application must scale based on incoming traffic, and it should have zero downtime during updates. Which of the following are crucial considerations to achieve this?
- A
Use a Deployment resource with a rolling update strategy.
- B
Configure Horizontal Pod Autoscaler (HPA) to scale pods based on CPU utilization.
- C
Deploy the application using a StatefulSet to ensure the state is preserved across pods.
- D
Enable liveness and readiness probes in the Pod specification.
- E
Manually scale the number of replicas to handle traffic spikes.
Show answer and explanation
Correct answers: A, B, D
Explanation
To deploy a stateless application on GKE with zero downtime and scalability, you should use a Deployment resource with a rolling update strategy to handle seamless updates. Implementing Horizontal Pod Autoscaler (HPA) ensures automatic scaling based on traffic, while liveness and readiness probes maintain application health and readiness. StatefulSet and manual scaling are not suitable for this scenario as they do not align with the stateless and automated scaling requirements.
- A. Correct.
Using a Deployment resource with a rolling update strategy ensures zero downtime during updates by gradually replacing old pods with new ones.
- B. Correct.
Configuring Horizontal Pod Autoscaler (HPA) allows the application to scale pods automatically based on metrics like CPU utilization, which is essential for handling variable traffic.
- C. Incorrect.
StatefulSet is designed for stateful applications with unique identities for each pod. Since the application is stateless, a StatefulSet is not required.
- D. Correct.
Liveness and readiness probes ensure that the application is functioning correctly and is ready to serve traffic, which is critical for seamless operation and zero downtime.
- E. Incorrect.
Manually scaling the number of replicas is not efficient for handling traffic spikes as it lacks automation and may lead to delayed responses during sudden load changes.