In which scenario is serverless computing most advantageous?

• A. Long-running, stateful workloads requiring low cold-start latency
• B. Event-driven workloads with variable traffic
• C. Predictable, constant high CPU usage
• D. Applications requiring persistent connections

Answer: (B)

Serverless computing is most advantageous when the workload is event-driven and traffic is variable because it auto-scales with demand and you pay only for actual execution time. This means as events spike, the platform runs more function instances to handle them, and as activity drops, it scales down toward zero, eliminating idle costs and the need to over-provision. It also simplifies operations since there’s no server provisioning or maintenance.

In contrast, long-running, stateful workloads with low cold-start latency can suffer from startup delays and managing state across ephemeral executions. Predictable, constant high CPU usage isn’t ideal for serverless due to per-invocation costs and scaling behavior, and applications requiring persistent connections can hit limits and latency issues.