Design a Cloud-Native, Horizontally Scalable, Fault-Tolerant Web Application for Millions of Users

Context

You are asked to design, deploy, and operate a cloud-based web application that serves millions of users. Assume a major cloud provider, but keep the design cloud-agnostic with examples as needed. Focus on justifications, trade-offs, and operational excellence.

Requirements

1. Compare cloud service models and usage:
   • Contrast IaaS, PaaS, and serverless, and explain when to use each.
2. Compute layer selection:
   • Choose between VMs, containers, and functions as the primary compute. Justify your choice and mention where the others still fit.
3. Traffic management and statelessness:
   • Describe load balancing, autoscaling policies, and stateless service design.
4. Data layer design:
   • Select data stores for different workloads (relational, NoSQL, object storage, in-memory cache). Explain consistency, durability, and backup/restore strategies.
5. Network and edge:
   • Outline VPC, subnets, security groups, firewalls/WAF, NAT, CDN, and apply zero-trust principles.
6. Security and secrets:
   • Plan IAM, secrets management, encryption in transit/at rest, and key rotation.
7. Availability and resilience:
   • Ensure high availability and disaster recovery (multi-AZ/region) with clear RTO/RPO targets.
8. Observability and cost:
   • Implement metrics, logs, traces, SLOs, and cost governance controls.
9. Delivery and operations:
   • Describe CI/CD, Infrastructure as Code (IaC), blue/green and canary releases, and rollback strategies.