阅读中文版本

Build a Minimal MQTT Broker

Minimal MQTT Broker — a complete walkthrough of publish-subscribe messaging, covering topic trie, subscription matcher, retained store, and testing.

mqtt_broker_minimal/
├── src/index.ts
├── src/topic-trie.ts
├── src/subscription-matcher.ts
├── tests/mqtt_broker_minimal.test.ts
└── package.json

// Topic trie: publish-subscribe messaging core
export function topictrie(input: TopictrieInput): TopictrieResult {
  // Build the topic trie that anchors publish-subscribe messaging and defines the core data shape.
  return processTopictrie(input);
}

// Subscription matcher — handles publish-subscribe messaging Add the subscription matcher so
export class Subscriptionmatcher {
  constructor(private opts: SubscriptionmatcherOptions) {}
  async run(ctx: Context): Promise<SubscriptionmatcherOutput> {
    return this.execute(ctx);
  }
}

// Retained store: publish-subscribe messaging stage
export const retainedstore = (next: Next) => async (req: Request) => {
  const out = await transform(req);
  return next(out);
};

// QoS handshake for publish-subscribe messaging
export async function qoshandshake(stream: AsyncIterable<Chunk>): Promise<Aggregate> {
  let acc = init();
  for await (const c of stream) acc = merge(acc, c);
  return acc;
}

// Session resume: publish-subscribe messaging finalization
export function sessionresume(state: State): Result {
  validate(state);
  return finalize(state);
}

Project overview

This project builds a publish-subscribe messaging system whose reason for existing is matching topic filters with wildcards and delivering retained messages to late subscribers. The temptation in publish-subscribe messaging is to bolt the interesting part onto a working prototype and discover the hard cases in production; this build inverts that order and treats a subscriber with a wildcard that matches a high-volume topic and is overwhelmed by the fanout as the first thing designed, not the last.

You will implement Topic trie, Subscription matcher, Retained store, QoS handshake, and Session resume, each as a module with one responsibility and a test that proves it in isolation. The publish-subscribe messaging pipeline is then the wiring that composes them, and the integration test is where the composition is shown to survive a subscriber with a wildcard that matches a high-volume topic and is overwhelmed by the fanout.

The walkthrough is written for a developer who has used a publish-subscribe messaging system but has not had to make one honest. It favors explicit data shapes over clever abstractions, and it names the failure each design choice prevents rather than presenting the choice as a style preference.

Each phase ends with something observable — a topic trie that returns the right shape, a subscription matcher that rejects bad input, an integration test that goes green — so you can stop at any point and still have a publish-subscribe messaging system that works for the cases covered so far.

What you will build

The finished publish-subscribe messaging system exposes five capabilities, each with a focused test:

• Topic trie
• Subscription matcher
• Retained store
• QoS handshake
• Session resume

Topic trie is the foundation that every later stage assumes; Subscription matcher is where publish-subscribe messaging correctness lives; Retained store handles a subscriber with a wildcard that matches a high-volume topic and is overwhelmed by the fanout; QoS handshake makes the system observable; and Session resume is the graceful-degradation layer that keeps overload from becoming corruption.

The build keeps the publish-subscribe messaging logic in ordinary modules you can read in one sitting, and pulls in dependencies only for the parts that are genuinely tedious to hand-roll. That keeps the diff between the tutorial and a real codebase small, and keeps the failure modes visible in code rather than hidden behind a framework.

Architecture and data model

The system is organized around Topic trie, Subscription matcher, and Retained store, with QoS handshake tying them together. Data flows as a trie of topic levels where leaf nodes hold subscriber sets, plus a retained-message map keyed by exact topic.

Each component is a module with a pure interface where practical: the same input produces the same output, and side effects are pushed to the edges. That discipline is what makes the phases independently testable, because a test for Retained store does not need to stand up Topic trie and Subscription matcher to exercise the interesting branch.

The data model is deliberately explicit. A trie of topic levels where leaf nodes hold subscriber sets, plus a retained-message map keyed by exact topic Keeping the shape written down — in a schema, a type, or a fixture — means a change to the shape is a visible diff rather than a silent runtime shift, and it means the integration test can assert against a concrete shape instead of guessing.

Configuration and code are kept apart: the configuration says what publish-subscribe messaging should do, the modules say how. A reviewer can change a threshold or a policy without reading the implementation, which keeps the risky changes small and lets the diff stand as the review.

The failure path through Retained store is designed before the happy path is finished, because a subscriber with a wildcard that matches a high-volume topic and is overwhelmed by the fanout is the case that distinguishes a publish-subscribe messaging system from a demo. Every component declares what it does when its inputs are wrong, and the integration test asserts each of those declarations.

Prerequisites and dependencies

Have these ready before you install anything, so the publish-subscribe messaging setup is not blocked by a missing tool:

• Node.js 20 LTS or later
• Familiarity with publish-subscribe messaging fundamentals
• A local environment to run the test suite

The runtime dependencies are chosen to keep the publish-subscribe messaging datapath legible:

• aedes (^0.54.0) — MQTT protocol handling
• fastify (^5.0.0) — Broker control API

# setup for mqtt-broker-minimal
pnpm init
pnpm add aedes fastify
pnpm add -D typescript tsx vitest

Versions are pinned to a major range so the publish-subscribe messaging install is reproducible; commit the lockfile. The topic trie reads its operational parameters from the environment and validates them at startup, so a missing value fails the boot rather than the first publish-subscribe messaging request.

Directory structure

Keep each publish-subscribe messaging component in its own file so it can be tested without starting the others:

mqtt_broker_minimal/
├── src/index.ts
├── src/topic-trie.ts
├── src/subscription-matcher.ts
├── tests/mqtt_broker_minimal.test.ts
└── package.json

Phase 1: Topic trie

Build the topic trie that anchors publish-subscribe messaging and defines the core data shape.

// Topic trie: publish-subscribe messaging core
export function topictrie(input: TopictrieInput): TopictrieResult {
  // Build the topic trie that anchors publish-subscribe messaging and defines the core data shape.
  return processTopictrie(input);
}

After this phase the topic trie works in isolation and is covered by a unit test. The topic trie defines the data shape that every later stage assumes, so getting it right first means the later phases compile against a stable contract.

Treat the topic trie as the single source of truth for the publish-subscribe messaging data shape. Validate inputs at the boundary and reject early, so a malformed value never reaches Subscription matcher or Retained store where its effect would be harder to localize.

Phase 2: Subscription matcher

Add the subscription matcher so publish-subscribe messaging can matching topic filters with correctly.

// Subscription matcher — handles publish-subscribe messaging Add the subscription matcher so
export class Subscriptionmatcher {
  constructor(private opts: SubscriptionmatcherOptions) {}
  async run(ctx: Context): Promise<SubscriptionmatcherOutput> {
    return this.execute(ctx);
  }
}

After this phase the subscription matcher works in isolation and is covered by a unit test. This is where publish-subscribe messaging correctness lives: the subscription matcher is what makes matching topic filters with wildcards and delivering retained messages to late subscribers true, and its test is the proof.

Keep the subscription matcher pure where you can. A function that takes an input and returns an output is trivial to test and impossible to mock wrong; a class with hidden state is the opposite. Push side effects to the caller and keep the subscription matcher about the transformation.

Phase 3: Retained store

Wire the retained store to handle a subscriber with a without losing data.

// Retained store: publish-subscribe messaging stage
export const retainedstore = (next: Next) => async (req: Request) => {
  const out = await transform(req);
  return next(out);
};

After this phase the retained store works in isolation and is covered by a unit test. The retained store is the difference between a publish-subscribe messaging system that works on the happy path and one that survives a subscriber with a wildcard that matches a high-volume topic and is overwhelmed by the fanout.

Design the failure before the success. Decide what retained store does when a subscriber with a wildcard that matches a high-volume topic and is overwhelmed by the fanout: reject, retry, shed, or degrade — and write that test first. A publish-subscribe messaging system whose failure behavior is unspecified will pick one at random under pressure, and it is usually the wrong one.

Phase 4: QoS handshake

Integrate the qos handshake to make publish-subscribe messaging observable and testable end to end.

// QoS handshake for publish-subscribe messaging
export async function qoshandshake(stream: AsyncIterable<Chunk>): Promise<Aggregate> {
  let acc = init();
  for await (const c of stream) acc = merge(acc, c);
  return acc;
}

After this phase the qos handshake works in isolation and is covered by a unit test. Observability is not a finishing touch; it is how you know the publish-subscribe messaging system is doing what it claims in production.

Emit one structured event per meaningful action, with enough context to localize a failure without a second round-trip. The qos handshake is what turns “something is slow” into “the subscription matcher is slow for inputs matching this shape,” which is the difference between an alert and an investigation.

Phase 5: Session resume

Finalize the session resume so the publish-subscribe messaging system degrades gracefully under overload.

// Session resume: publish-subscribe messaging finalization
export function sessionresume(state: State): Result {
  validate(state);
  return finalize(state);
}

After this phase the session resume works in isolation and is covered by a unit test. The session resume is what keeps overload from becoming corruption: when the publish-subscribe messaging system is asked for more than it can do, it degrades in a defined way instead of failing in an undefined one.

Pick a degradation strategy for session resume — shed, queue, or backpressure — and make it explicit in code and in a test. Leaving the behavior to chance is how a publish-subscribe messaging traffic spike becomes a data loss incident.

Testing

Each phase has a focused test: topic trie, subscription matcher, and retained store are exercised in isolation, then an integration test drives the full publish-subscribe messaging pipeline and asserts the a subscriber with a case is handled.

Start with unit tests that pin each component’s contract: the topic trie returns the documented shape, the subscription matcher transforms correctly, and the retained store fails in the documented way. These tests are fast and they localize regressions to the component that broke.

Then add an integration test that composes the components and drives a realistic publish-subscribe messaging scenario end to end. Assert the happy path first, then deliberately trigger a subscriber with a wildcard that matches a high-volume topic and is overwhelmed by the fanout and assert the system degrades rather than corrupts. Because the components are pure where practical, the integration test stays deterministic rather than flaky.

A publish-subscribe messaging suite must also prove the negative: bad input is rejected at the boundary, overload is shed rather than queued indefinitely, and a partial failure does not leave the system in an inconsistent state. These are the tests that catch the regressions a happy-path suite never would.

The entire publish-subscribe messaging suite runs with no external service; every dependency is either in-process or stubbed, so every case is reproducible in CI. That is what lets the retained store failure-mode assertions live as tests rather than as a manual checklist that drifts out of date.

# run the publish-subscribe messaging test suite
pnpm vitest run

Deployment and operations

Run at least two publish-subscribe messaging instances behind a load balancer and share state in Redis. Set resource limits so publish-subscribe messaging overload surfaces as backpressure rather than failure.

Run at least two stateless instances behind a load balancer, and move any shared state into a shared store so the publish-subscribe messaging behavior is global rather than per instance. Size limits so the system sheds load before it runs out of memory, because a publish-subscribe messaging system that OOMs under load fails in a way that is hard to reason about.

Drain on deploy: stop accepting new work, let in-flight publish-subscribe messaging operations finish, then exit. Because the system holds no state that cannot be reconstructed, a rollout is a matter of cycling instances one at a time while watching the error rate.

When a dependency is unavailable, decide explicitly whether the publish-subscribe messaging system fails closed or degrades. Leaving that choice to chance is how a dependency outage becomes a silent data loss, so make the choice visible in code and in a test.

Security considerations

Validate all publish-subscribe messaging inputs at the boundary, reject oversized payloads before buffering them, and keep secrets in the environment rather than the repository.

The publish-subscribe messaging system is a choke point, so it is the right place to reject malformed input before it reaches subscription matcher or retained store. Validate the shape and the size at the edge, and let the inner components assume what they receive is well-formed.

Never log secrets or full payloads; log enough to localize a failure without recording data that should not be on disk. Treat any external call as untrusted and validate its response before using it, because a publish-subscribe messaging system that trusts its inputs blindly is one malformed response away from corruption.

Keep dependencies pinned and run an audit in CI. A transitive dependency with a known vulnerability is still a vulnerability in your publish-subscribe messaging system, and the lockfile is what makes the audit meaningful.

Observability

Emit one structured event per publish-subscribe messaging operation with the inputs’ shape, the outcome, and a correlation id. Track error rate and latency per component and alert on a step change rather than a single spike.

The correlation id is the thread that ties a user-visible symptom to the publish-subscribe messaging decision and to the component log entry for the same operation. Without it, an investigation starts from a timestamp and a guess.

Watch the retained store failure rate separately from the overall error rate. A jump in retained store failures usually means a subscriber with a wildcard that matches a high-volume topic and is overwhelmed by the fanout, not a bug in subscription matcher, and the two have different owners and different fixes.

Chart the share of publish-subscribe messaging operations that shed, queue, or succeed. A change in the mix points at the cause faster than an aggregate latency can, because each class has a different root.

Troubleshooting

A subscriber with a wildcard that matches a high-volume topic and is overwhelmed by the fanout. Reproduce with a minimal fixture, then add a guard test so the publish-subscribe messaging system rejects the condition loudly.

Performance regresses under publish-subscribe messaging load. Profile with the sampling analyzer, locate the hot path in subscription matcher, and add a targeted benchmark to lock the regression.

Keep the publish-subscribe messaging dashboards separate from dependency dashboards. When an operation fails, the correlation id lets you decide in seconds whether the fault is topic trie, subscription matcher, or retained store, instead of guessing from a timeout.

Limitations and next steps

• Single-region; multi-region replication is out of scope.
• The topic trie is in-process and not horizontally shared by default.
• No auth layer; assume deployment behind an edge that enforces identity.

The publish-subscribe messaging system is single-region by design; extending to multi-region means making topic trie shareable across regions, which is a real change to the consistency model rather than a config flag. The retained store is in-process, which is correct for one instance but must move to a shared store before scaling out.

There is no authentication or authorization layer; the system assumes it is deployed behind an edge that enforces identity. Adding one is straightforward but out of scope for this build, and doing it here would distract from the publish-subscribe messaging correctness that is the point.

The natural next step, once the publish-subscribe messaging pipeline is trusted in production, is to add a short-TTL cache for idempotent reads and a richer qos handshake that exports histograms rather than counters. Both are additive and neither changes the contracts this build establishes.