Choose an Execution Model for the Storefront Tier
Choosing an execution model weighs virtual machines, managed app platforms, containers, and serverless functions in terms of control, operational burden, and scaling. This interactive lesson compares all four for a storefront API and explains why Amazon ECS on AWS Fargate is the preferred choice. Covers: Ownership boundaries, when not to use each model, provider mappings such as EC2, App Service, Cloud Run, EKS, and Lambda, and two-minute decision answers. Prerequisites: VM, container, and serverless basics.
We'll cover the following...
In a system-design interview, you’re asked where to run the storefront’s application programming interface (API) tier. It receives spiky seasonal traffic through an ingress layer, serves latency-sensitive synchronous requests, and starts background work without redesigning its private managed data dependencies. The interviewer is scoring whether we connect those requirements to operational ownership and failure behavior, not whether we can recite service names.
Let’s assume frequent deployments, a small on-call team, and compatibility needs that may include a conventional web framework, background workers, or persistent connection patterns. We will evaluate control and compatibility, operational ownership, scaling fit, and startup latency. Those dimensions make virtual machines, managed platforms, containers, and functions plausible alternatives rather than interchangeable answers.
Inspect how these dimensions separate the four execution models for the ...