Serverless Building Blocks

Serverless architecture is rarely just “a function that runs.” A real system is usually made of several managed layers working together: entry points, event sources, compute, durable state, messaging, identity, and workflow or notification services. Chapter 2 builds that vocabulary so later pattern discussions do not collapse into vague references to “the platform.”

Read the lessons in order. The first focuses on functions as the core compute primitive. The second explains triggers and reactive entry points. The third shows why serverless usually means many managed backend services working together. The fourth explains why stateless compute changes how the system handles persistence, progress, and recovery.

If a serverless design still looks like “one function plus some cloud stuff,” this chapter is where the model becomes more precise.

In this section

• Functions as the Core Compute Primitive
  Explain function-based compute as a short-lived, event-triggered execution model. Describe startup, invocation, scaling, execution limits, and what functions are good at.
• Events, Triggers, and Reactive Entry Points
  Describe the ways functions are triggered: HTTP, queues, topics, streams, schedules, object changes, authentication events, and platform hooks. This section should show that serverless is naturally event-driven.
• Backend Services and Managed Building Blocks
  Introduce databases, object storage, authentication, secrets, messaging, notifications, and workflow services as core parts of the serverless toolbox. Explain why serverless usually means "many managed services working together."
• Stateless Compute, External State
  Explain why serverless functions should usually be treated as ephemeral and stateless, with durable state moved into external stores and managed systems.