Skip to main content

Chapter 54: Dapr Actors & Workflows

Stateless pub/sub and service invocation handle many patterns. But agents often need state—conversation history, user preferences, task progress. And long-running operations—multi-step workflows that survive restarts, with retries and compensation logic.

Dapr Actors provide virtual, stateful entities that encapsulate agent state. Dapr Workflows provide durable orchestration for long-running processes. Together, they enable sophisticated agent patterns like persistent conversations, scheduled reminders, and multi-step task automation.

What You'll Learn​

By the end of this chapter, you'll be able to:

  • Understand the Actor model: Virtual actors, state encapsulation, and turn-based concurrency
  • Implement Dapr Actors: Define actors for agent state (conversations, user sessions)
  • Manage actor lifecycle: Activation, deactivation, and reminders
  • Use actor timers & reminders: Schedule future operations
  • Design Dapr Workflows: Durable, restartable orchestration
  • Implement workflow patterns: Sequential, parallel, and saga patterns
  • Handle failures: Compensation, retries, and error handling in workflows
  • Combine actors & workflows: Actors for state, workflows for orchestration

Chapter Structure​

  1. The Actor Model — Concepts, benefits, and when to use actors
  2. Dapr Actors Fundamentals — Defining actors, state, and methods
  3. Actor State Management — Persisting and retrieving actor state
  4. Timers & Reminders — Scheduling future actor invocations
  5. Dapr Workflows Overview — Durable orchestration concepts
  6. Workflow Implementation — Defining workflows, activities, and child workflows
  7. Workflow Patterns — Sequential, fan-out/fan-in, and sagas
  8. Capstone: Stateful Agent with Reminders — Build an agent with persistent conversations and scheduled task reminders

Prerequisites​

  • Chapter 53: Dapr Fundamentals (building blocks, components)
  • Your agent service using Dapr pub/sub and state

Looking Ahead​

You now have a stateful, event-driven agent on Kubernetes. Chapter 55 automates deployments with CI/CD and GitOps, ensuring every code change flows to production safely.