Learning track · Roadmap
Streaming & Event-Driven Architecture — Learning Track (Part 4)
Published
Your map for Part 4 of the series. Parts 1–3 covered distributed systems, transactions & isolation, and storage engines. This part is about how data moves and is processed as it arrives.
- Mission: design event-driven systems with judgment — the log as a unifying abstraction, Kafka, event sourcing, CQRS, stream processing.
- Format: log/dataflow/event-time diagrams + concrete examples. Delivered as a lean EPUB for Kindle (glossary kept in the
.md, out of the book) plus markdown source here. - Level: builds on Parts 1–3; each lesson adds senior-level depth (real-system behaviour).
Where you are now
| Status | All 9 lessons built ✅ — read as one comprehensive page + deep-dive supplements |
| Read it | book-4-streaming-event-driven-fundamentals.epub (send to Kindle) · or streaming-event-driven-fundamentals.md |
| How to work it | Read in order; do each self-check from memory before peeking |
| Deep dive | watermarks-deep-dive.epub — a go-deeper supplement to Lesson 8 (perfect vs heuristic · the idle-partition stall · triggers & late firings) |
| Deep dive | exactly-once-kafka-deep-dive.epub — a go-deeper supplement to Lesson 3 (idempotent producer PID+seq · atomic offset commit · read_committed/LSO · zombie fencing & the epoch) |
The 9-lesson path
| # | Lesson | The single win | Status |
|---|---|---|---|
| 1 | The Log: An Append-Only Source of Truth | Why the log is the unifying abstraction | ✅ Built |
| 2 | Kafka: The Distributed Log | Topics, partitions, offsets, consumer groups | ✅ Built |
| 3 | Delivery Guarantees in Streams | The offset-commit point decides everything | ✅ Built |
| 4 | Events vs Commands vs State | Thin notification vs fat state transfer | ✅ Built |
| 5 | Event Sourcing | Events as the source of truth | ✅ Built |
| 6 | CQRS | One write model, many read models | ✅ Built |
| 7 | Stream Processing | Stateless vs stateful dataflow | ✅ Built |
| 8 | Time, Windows & Watermarks | Event time, late data, completeness vs latency | ✅ Built |
| 9 | Building It Right | Patterns & pitfalls (the checklist) | ✅ Built |
How every lesson is built: prose → a diagram → a self-check → an expert corner.
Progress checklist
- Lesson 1 — The Log
- Lesson 2 — Kafka
- Lesson 3 — Delivery Guarantees
- Lesson 4 — Events vs Commands vs State
- Lesson 5 — Event Sourcing
- Lesson 6 — CQRS
- Lesson 7 — Stream Processing
- Lesson 8 — Time, Windows & Watermarks
- Lesson 9 — Building It Right
Tick each box as you finish its self-check; tell me where you want to go deeper.
Files in this folder
README.md ← index + roadmap + tracker
streaming-event-driven-fundamentals.md ← full source (includes the glossary)
book-4-streaming-event-driven-fundamentals.epub ← lean Kindle build (glossary excluded)
diagrams/
00-cover.svg / .png ← series cover (Part 4)
01-the-log.svg / .png ← Lesson 1
02-kafka-partitions.svg / .png ← Lesson 2
03-offset-commit.svg / .png ← Lesson 3
04-event-types.svg / .png ← Lesson 4
05-event-sourcing.svg / .png ← Lesson 5
06-cqrs.svg / .png ← Lesson 6
07-stream-processing.svg / .png ← Lesson 7
08-windows-watermark.svg / .png ← Lesson 8
09-eventdriven-patterns.svg / .png ← Lesson 9Next in the series
Part 5 — Applied Systems Design (the capstone): design a rate limiter, a news feed, a distributed cache, a message queue, a URL shortener — applying all four prior books end-to-end. This is the final book of the series.