Skip to content

02-system-architecture-app-worker-engine.md

Latest commit

 

History

History
52 lines (31 loc) · 2.4 KB

02-system-architecture-app-worker-engine.md

File metadata and controls

52 lines (31 loc) · 2.4 KB
layout default
title Chapter 2: System Architecture: App, Worker, Engine
nav_order 2
parent Activepieces Tutorial

Chapter 2: System Architecture: App, Worker, Engine

Welcome to Chapter 2: System Architecture: App, Worker, Engine. In this part of Activepieces Tutorial: Open-Source Automation, Pieces, and AI-Ready Workflow Operations, you will build an intuitive mental model first, then move into concrete implementation details and practical production tradeoffs.

This chapter explains the execution architecture that powers scale and reliability.

Learning Goals

  • understand responsibilities of app, worker, engine, and UI layers
  • map queue/database dependencies to execution behavior
  • identify likely bottlenecks under load spikes
  • plan scaling steps without breaking flow reliability

Architecture Summary

Activepieces separates concerns:

  • app handles APIs, validation, and orchestration entrypoints
  • worker pulls jobs and executes flows via the engine
  • engine parses flow JSON and runs execution logic
  • Postgres + Redis back persistent state and queueing behavior

Scaling Heuristic

Start by scaling workers for execution pressure, then app replicas for ingress/API load, while keeping Redis/Postgres availability and capacity visible through monitoring.

Source References

Summary

You now understand the runtime surfaces that matter for production scaling decisions.

Next: Chapter 3: Flow Design, Versioning, and Debugging

Source Code Walkthrough

docker-compose.yml

The docker-compose.yml makes the app/worker/engine split concrete. The AP_CONTAINER_TYPE=APP and AP_CONTAINER_TYPE=WORKER environment variables confirm the two-process deployment model described in this chapter — a single image, different runtime roles. The worker service also shows replicas: 5, which reflects the horizontal scaling design of the execution layer.

For the broader monorepo package structure, the package.json workspace scripts (serve:backend, serve:worker, serve:engine) map directly to the architectural boundaries between the API, engine, and worker packages.