Correct Answer: C

Utilizing Google Kubernetes Engine (GKE) with cluster autoscaling enabled is the most effective strategy. It automatically scales the service based on demand, efficiently handling sudden spikes in user requests. GKE's cluster autoscaling adjusts the number of nodes in response to traffic changes, ensuring optimal resource allocation and application performance.

Why Not the Others?

• A: Deploying microservices on individual Compute Engine instances without autoscaling doesn't efficiently manage resources or varying workloads, leading to potential underutilization or performance issues during high demand.
• B: Implementing a custom cache eviction policy in Cloud CDN isn't feasible because Cloud CDN doesn't support custom cache eviction policies. Its policies are based on Cache-Control headers and default TTL settings, focusing on static content caching, not dynamic microservices traffic spikes.
• D: While a Cloud Load Balancer with global backend services can distribute traffic evenly, it lacks inherent autoscaling capabilities. Autoscaling must be enabled separately, making this approach insufficient alone for handling sudden traffic spikes efficiently.