Explain Serverless vs Containers.

Serverless runs event-driven functions on managed platforms, while containers package services and dependencies for portable, self-managed deployment. (#definition)

When to Use

• Serverless: Best for spiky workloads, lightweight APIs, webhooks, cron jobs, and ETL pipelines.
• Containers: Ideal for steady traffic apps, microservices, custom runtimes, long-lived connections, and stateful services.

Example

• Serverless: Uploading an image triggers a function to resize and save it.
• Containers: A product API with Redis and gRPC running in Kubernetes.

To master these trade-offs for interviews, check out Grokking System Design Fundamentals, Grokking the Coding Interview, and Mock Interviews with ex-FAANG engineers.

Why Is It Important

The right choice affects cost, latency, scalability, and maintainability. Many real-world systems use a hybrid model.

Interview Tips

• Start with a crisp one-liner definition.
• Compare scaling, startup latency, control, cost model, and ops.
• Suggest hybrid use cases and mitigations like provisioned concurrency (serverless) or auto-scaling (containers).

Trade-offs

• Serverless gains: no ops, scale-to-zero, pay-per-use.
• Serverless drawbacks: cold starts, vendor lock-in, timeout limits.
• Containers gains: portability, full runtime control.
• Containers drawbacks: cluster management, patching, idle costs.

Pitfalls

• Using serverless for chatty, high-throughput services.
• Packing too many processes in one container.
• Ignoring concurrency, idempotency, and observability.
• Overlooking cloud limits (timeouts, networking, region availability).