Answer: Create a global load balancer with managed instance groups and autoscaling policies. Use non-preemptible Compute Engine instances.

The Mountkirk Games case study emphasizes business requirements such as improving uptime (where downtime results in loss of players), ensuring game stability and speed (as per KPIs demanded by investors), and reducing latency for global customers. Preemptible instances, while cost-effective, can be terminated with minimal notice (up to 30 seconds), leading to potential disruptions in gaming sessions and violating the uptime and stability requirements. A global load balancer with managed instance groups and autoscaling policies ensures high availability, scalability, and low latency across regions, aligning with the need for a reliable, scalable backend. Non-preemptible instances provide the necessary reliability, avoiding interruptions that could degrade user experience. Community discussion strongly supports option D, with high upvotes (e.g., 50 upvotes for a comment highlighting KPI stability and non-preemptible instances) and consensus that preemptible instances are unsuitable for fault-intolerant applications like gaming. Options A and B use network load balancers, which lack global distribution and may not optimize latency. Option C uses preemptible instances, which risk instability and data loss, contradicting the case study's focus on uptime and KPIs.

Author: LeetQuiz Editorial Team