阅读中文版本

Build a WAL Replication Manager

WAL Replication Manager — a complete walkthrough of write-ahead log replication, covering log appender, follower shipper, checkpoint store, and testing.

wal_replication_manager/
├── src/index.ts
├── src/log-appender.ts
├── src/follower-shipper.ts
├── tests/wal_replication_manager.test.ts
└── package.json

// Log appender: write-ahead log replication core
export function logappender(input: LogappenderInput): LogappenderResult {
  // Build the log appender that anchors write-ahead log replication and defines the core data shape.
  return processLogappender(input);
}

// Follower shipper — handles write-ahead log replication Add the follower shipper so
export class Followershipper {
  constructor(private opts: FollowershipperOptions) {}
  async run(ctx: Context): Promise<FollowershipperOutput> {
    return this.execute(ctx);
  }
}

// Checkpoint store: write-ahead log replication stage
export const checkpointstore = (next: Next) => async (req: Request) => {
  const out = await transform(req);
  return next(out);
};

// Gap detector for write-ahead log replication
export async function gapdetector(stream: AsyncIterable<Chunk>): Promise<Aggregate> {
  let acc = init();
  for await (const c of stream) acc = merge(acc, c);
  return acc;
}

// Catch-up replay: write-ahead log replication finalization
export function catchupreplay(state: State): Result {
  validate(state);
  return finalize(state);
}

Project overview

This project builds a write-ahead log replication system whose reason for existing is shipping an ordered log to followers so they converge without gaps after a leader failover. The temptation in write-ahead log replication 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 follower that falls behind the log retention window and must request a full snapshot instead of replaying entries as the first thing designed, not the last.

You will implement Log appender, Follower shipper, Checkpoint store, Gap detector, and Catch-up replay, each as a module with one responsibility and a test that proves it in isolation. The write-ahead log replication pipeline is then the wiring that composes them, and the integration test is where the composition is shown to survive a follower that falls behind the log retention window and must request a full snapshot instead of replaying entries.

The walkthrough is written for a developer who has used a write-ahead log replication 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 log appender that returns the right shape, a follower shipper that rejects bad input, an integration test that goes green — so you can stop at any point and still have a write-ahead log replication system that works for the cases covered so far.

What you will build

The finished write-ahead log replication system exposes five capabilities, each with a focused test:

• Log appender
• Follower shipper
• Checkpoint store
• Gap detector
• Catch-up replay

Log appender is the foundation that every later stage assumes; Follower shipper is where write-ahead log replication correctness lives; Checkpoint store handles a follower that falls behind the log retention window and must request a full snapshot instead of replaying entries; Gap detector makes the system observable; and Catch-up replay is the graceful-degradation layer that keeps overload from becoming corruption.

The build keeps the write-ahead log replication 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 Log appender, Follower shipper, and Checkpoint store, with Gap detector tying them together. Data flows as a monotonically increasing sequence of log entries replicated to followers, with a checkpoint marking the durable tail.

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 Checkpoint store does not need to stand up Log appender and Follower shipper to exercise the interesting branch.

The data model is deliberately explicit. A monotonically increasing sequence of log entries replicated to followers, with a checkpoint marking the durable tail 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 write-ahead log replication 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 Checkpoint store is designed before the happy path is finished, because a follower that falls behind the log retention window and must request a full snapshot instead of replaying entries is the case that distinguishes a write-ahead log replication 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 write-ahead log replication setup is not blocked by a missing tool:

• Node.js 20 LTS or later
• Familiarity with write-ahead log replication fundamentals
• A local environment to run the test suite

The runtime dependencies are chosen to keep the write-ahead log replication datapath legible:

• fastify (^5.0.0) — Replication control plane
• lmdb (^3.2.0) — Local log store

# setup for wal-replication-manager
pnpm init
pnpm add fastify lmdb
pnpm add -D typescript tsx vitest

Versions are pinned to a major range so the write-ahead log replication install is reproducible; commit the lockfile. The log appender reads its operational parameters from the environment and validates them at startup, so a missing value fails the boot rather than the first write-ahead log replication request.

Directory structure

Keep each write-ahead log replication component in its own file so it can be tested without starting the others:

wal_replication_manager/
├── src/index.ts
├── src/log-appender.ts
├── src/follower-shipper.ts
├── tests/wal_replication_manager.test.ts
└── package.json

Phase 1: Log appender

Build the log appender that anchors write-ahead log replication and defines the core data shape.

// Log appender: write-ahead log replication core
export function logappender(input: LogappenderInput): LogappenderResult {
  // Build the log appender that anchors write-ahead log replication and defines the core data shape.
  return processLogappender(input);
}

After this phase the log appender works in isolation and is covered by a unit test. The log appender 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 log appender as the single source of truth for the write-ahead log replication data shape. Validate inputs at the boundary and reject early, so a malformed value never reaches Follower shipper or Checkpoint store where its effect would be harder to localize.

Phase 2: Follower shipper

Add the follower shipper so write-ahead log replication can shipping an ordered log correctly.

// Follower shipper — handles write-ahead log replication Add the follower shipper so
export class Followershipper {
  constructor(private opts: FollowershipperOptions) {}
  async run(ctx: Context): Promise<FollowershipperOutput> {
    return this.execute(ctx);
  }
}

After this phase the follower shipper works in isolation and is covered by a unit test. This is where write-ahead log replication correctness lives: the follower shipper is what makes shipping an ordered log to followers so they converge without gaps after a leader failover true, and its test is the proof.

Keep the follower shipper 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 follower shipper about the transformation.

Phase 3: Checkpoint store

Wire the checkpoint store to handle a follower that falls without losing data.

// Checkpoint store: write-ahead log replication stage
export const checkpointstore = (next: Next) => async (req: Request) => {
  const out = await transform(req);
  return next(out);
};

After this phase the checkpoint store works in isolation and is covered by a unit test. The checkpoint store is the difference between a write-ahead log replication system that works on the happy path and one that survives a follower that falls behind the log retention window and must request a full snapshot instead of replaying entries.

Design the failure before the success. Decide what checkpoint store does when a follower that falls behind the log retention window and must request a full snapshot instead of replaying entries: reject, retry, shed, or degrade — and write that test first. A write-ahead log replication system whose failure behavior is unspecified will pick one at random under pressure, and it is usually the wrong one.

Phase 4: Gap detector

Integrate the gap detector to make write-ahead log replication observable and testable end to end.

// Gap detector for write-ahead log replication
export async function gapdetector(stream: AsyncIterable<Chunk>): Promise<Aggregate> {
  let acc = init();
  for await (const c of stream) acc = merge(acc, c);
  return acc;
}

After this phase the gap detector works in isolation and is covered by a unit test. Observability is not a finishing touch; it is how you know the write-ahead log replication 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 gap detector is what turns “something is slow” into “the follower shipper is slow for inputs matching this shape,” which is the difference between an alert and an investigation.

Phase 5: Catch-up replay

Finalize the catch-up replay so the write-ahead log replication system degrades gracefully under overload.

// Catch-up replay: write-ahead log replication finalization
export function catchupreplay(state: State): Result {
  validate(state);
  return finalize(state);
}

After this phase the catch-up replay works in isolation and is covered by a unit test. The catch-up replay is what keeps overload from becoming corruption: when the write-ahead log replication 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 catch-up replay — shed, queue, or backpressure — and make it explicit in code and in a test. Leaving the behavior to chance is how a write-ahead log replication traffic spike becomes a data loss incident.

Testing

Each phase has a focused test: log appender, follower shipper, and checkpoint store are exercised in isolation, then an integration test drives the full write-ahead log replication pipeline and asserts the a follower that falls case is handled.

Start with unit tests that pin each component’s contract: the log appender returns the documented shape, the follower shipper transforms correctly, and the checkpoint 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 write-ahead log replication scenario end to end. Assert the happy path first, then deliberately trigger a follower that falls behind the log retention window and must request a full snapshot instead of replaying entries and assert the system degrades rather than corrupts. Because the components are pure where practical, the integration test stays deterministic rather than flaky.

A write-ahead log replication 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 write-ahead log replication 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 checkpoint store failure-mode assertions live as tests rather than as a manual checklist that drifts out of date.

# run the write-ahead log replication test suite
pnpm vitest run

Deployment and operations

Run at least two write-ahead log replication instances behind a load balancer and share state in Redis. Set resource limits so write-ahead log replication 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 write-ahead log replication behavior is global rather than per instance. Size limits so the system sheds load before it runs out of memory, because a write-ahead log replication 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 write-ahead log replication 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 write-ahead log replication 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 write-ahead log replication inputs at the boundary, reject oversized payloads before buffering them, and keep secrets in the environment rather than the repository.

The write-ahead log replication system is a choke point, so it is the right place to reject malformed input before it reaches follower shipper or checkpoint 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 write-ahead log replication 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 write-ahead log replication system, and the lockfile is what makes the audit meaningful.

Observability

Emit one structured event per write-ahead log replication 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 write-ahead log replication decision and to the component log entry for the same operation. Without it, an investigation starts from a timestamp and a guess.

Watch the checkpoint store failure rate separately from the overall error rate. A jump in checkpoint store failures usually means a follower that falls behind the log retention window and must request a full snapshot instead of replaying entries, not a bug in follower shipper, and the two have different owners and different fixes.

Chart the share of write-ahead log replication 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 follower that falls behind the log retention window and must request a full snapshot instead of replaying entries. Reproduce with a minimal fixture, then add a guard test so the write-ahead log replication system rejects the condition loudly.

Performance regresses under write-ahead log replication load. Profile with the sampling analyzer, locate the hot path in follower shipper, and add a targeted benchmark to lock the regression.

Keep the write-ahead log replication dashboards separate from dependency dashboards. When an operation fails, the correlation id lets you decide in seconds whether the fault is log appender, follower shipper, or checkpoint store, instead of guessing from a timeout.

Limitations and next steps

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

The write-ahead log replication system is single-region by design; extending to multi-region means making log appender shareable across regions, which is a real change to the consistency model rather than a config flag. The checkpoint 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 write-ahead log replication correctness that is the point.

The natural next step, once the write-ahead log replication pipeline is trusted in production, is to add a short-TTL cache for idempotent reads and a richer gap detector that exports histograms rather than counters. Both are additive and neither changes the contracts this build establishes.