阅读中文版本

Build a Snapshot Backup Service

Snapshot Backup Service — a complete walkthrough of point-in-time backup, covering copy-on-write fork, chunk uploader, manifest writer, and testing.

snapshot_backup_service/
├── src/index.ts
├── src/copy-on-write-fork.ts
├── src/chunk-uploader.ts
├── tests/snapshot_backup_service.test.ts
└── package.json

// Copy-on-write fork: point-in-time backup core
export function copyonwritefork(input: CopyonwriteforkInput): CopyonwriteforkResult {
  // Build the copy-on-write fork that anchors point-in-time backup and defines the core data shape.
  return processCopyonwritefork(input);
}

// Chunk uploader — handles point-in-time backup Add the chunk uploader so
export class Chunkuploader {
  constructor(private opts: ChunkuploaderOptions) {}
  async run(ctx: Context): Promise<ChunkuploaderOutput> {
    return this.execute(ctx);
  }
}

// Manifest writer: point-in-time backup stage
export const manifestwriter = (next: Next) => async (req: Request) => {
  const out = await transform(req);
  return next(out);
};

// Restore planner for point-in-time backup
export async function restoreplanner(stream: AsyncIterable<Chunk>): Promise<Aggregate> {
  let acc = init();
  for await (const c of stream) acc = merge(acc, c);
  return acc;
}

// Integrity verifier: point-in-time backup finalization
export function integrityverifier(state: State): Result {
  validate(state);
  return finalize(state);
}

Project overview

This project builds a point-in-time backup system whose reason for existing is capturing a consistent snapshot of a mutable dataset while writers continue without blocking. The temptation in point-in-time backup 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 chunk upload that fails partway and leaves the manifest referencing a missing block as the first thing designed, not the last.

You will implement Copy-on-write fork, Chunk uploader, Manifest writer, Restore planner, and Integrity verifier, each as a module with one responsibility and a test that proves it in isolation. The point-in-time backup pipeline is then the wiring that composes them, and the integration test is where the composition is shown to survive a chunk upload that fails partway and leaves the manifest referencing a missing block.

The walkthrough is written for a developer who has used a point-in-time backup 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 copy-on-write fork that returns the right shape, a chunk uploader that rejects bad input, an integration test that goes green — so you can stop at any point and still have a point-in-time backup system that works for the cases covered so far.

What you will build

The finished point-in-time backup system exposes five capabilities, each with a focused test:

• Copy-on-write fork
• Chunk uploader
• Manifest writer
• Restore planner
• Integrity verifier

Copy-on-write fork is the foundation that every later stage assumes; Chunk uploader is where point-in-time backup correctness lives; Manifest writer handles a chunk upload that fails partway and leaves the manifest referencing a missing block; Restore planner makes the system observable; and Integrity verifier is the graceful-degradation layer that keeps overload from becoming corruption.

The build keeps the point-in-time backup 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 Copy-on-write fork, Chunk uploader, and Manifest writer, with Restore planner tying them together. Data flows as a manifest of content-addressed chunks plus a root pointer, reconstructable into a full dataset on restore.

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 Manifest writer does not need to stand up Copy-on-write fork and Chunk uploader to exercise the interesting branch.

The data model is deliberately explicit. A manifest of content-addressed chunks plus a root pointer, reconstructable into a full dataset on restore 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 point-in-time backup 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 Manifest writer is designed before the happy path is finished, because a chunk upload that fails partway and leaves the manifest referencing a missing block is the case that distinguishes a point-in-time backup 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 point-in-time backup setup is not blocked by a missing tool:

• Node.js 20 LTS or later
• Familiarity with point-in-time backup fundamentals
• A local environment to run the test suite

The runtime dependencies are chosen to keep the point-in-time backup datapath legible:

• express (^4.21.0) — Backup API
• blake3 (^0.2.0) — Chunk content addressing
• undici (^6.19.0) — Object store uploads

# setup for snapshot-backup-service
pnpm init
pnpm add express blake3 undici
pnpm add -D typescript tsx vitest

Versions are pinned to a major range so the point-in-time backup install is reproducible; commit the lockfile. The copy-on-write fork reads its operational parameters from the environment and validates them at startup, so a missing value fails the boot rather than the first point-in-time backup request.

Directory structure

Keep each point-in-time backup component in its own file so it can be tested without starting the others:

snapshot_backup_service/
├── src/index.ts
├── src/copy-on-write-fork.ts
├── src/chunk-uploader.ts
├── tests/snapshot_backup_service.test.ts
└── package.json

Phase 1: Copy-on-write fork

Build the copy-on-write fork that anchors point-in-time backup and defines the core data shape.

// Copy-on-write fork: point-in-time backup core
export function copyonwritefork(input: CopyonwriteforkInput): CopyonwriteforkResult {
  // Build the copy-on-write fork that anchors point-in-time backup and defines the core data shape.
  return processCopyonwritefork(input);
}

After this phase the copy-on-write fork works in isolation and is covered by a unit test. The copy-on-write fork 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 copy-on-write fork as the single source of truth for the point-in-time backup data shape. Validate inputs at the boundary and reject early, so a malformed value never reaches Chunk uploader or Manifest writer where its effect would be harder to localize.

Phase 2: Chunk uploader

Add the chunk uploader so point-in-time backup can capturing a consistent snapshot correctly.

// Chunk uploader — handles point-in-time backup Add the chunk uploader so
export class Chunkuploader {
  constructor(private opts: ChunkuploaderOptions) {}
  async run(ctx: Context): Promise<ChunkuploaderOutput> {
    return this.execute(ctx);
  }
}

After this phase the chunk uploader works in isolation and is covered by a unit test. This is where point-in-time backup correctness lives: the chunk uploader is what makes capturing a consistent snapshot of a mutable dataset while writers continue without blocking true, and its test is the proof.

Keep the chunk uploader 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 chunk uploader about the transformation.

Phase 3: Manifest writer

Wire the manifest writer to handle a chunk upload that without losing data.

// Manifest writer: point-in-time backup stage
export const manifestwriter = (next: Next) => async (req: Request) => {
  const out = await transform(req);
  return next(out);
};

After this phase the manifest writer works in isolation and is covered by a unit test. The manifest writer is the difference between a point-in-time backup system that works on the happy path and one that survives a chunk upload that fails partway and leaves the manifest referencing a missing block.

Design the failure before the success. Decide what manifest writer does when a chunk upload that fails partway and leaves the manifest referencing a missing block: reject, retry, shed, or degrade — and write that test first. A point-in-time backup system whose failure behavior is unspecified will pick one at random under pressure, and it is usually the wrong one.

Phase 4: Restore planner

Integrate the restore planner to make point-in-time backup observable and testable end to end.

// Restore planner for point-in-time backup
export async function restoreplanner(stream: AsyncIterable<Chunk>): Promise<Aggregate> {
  let acc = init();
  for await (const c of stream) acc = merge(acc, c);
  return acc;
}

After this phase the restore planner works in isolation and is covered by a unit test. Observability is not a finishing touch; it is how you know the point-in-time backup 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 restore planner is what turns “something is slow” into “the chunk uploader is slow for inputs matching this shape,” which is the difference between an alert and an investigation.

Phase 5: Integrity verifier

Finalize the integrity verifier so the point-in-time backup system degrades gracefully under overload.

// Integrity verifier: point-in-time backup finalization
export function integrityverifier(state: State): Result {
  validate(state);
  return finalize(state);
}

After this phase the integrity verifier works in isolation and is covered by a unit test. The integrity verifier is what keeps overload from becoming corruption: when the point-in-time backup 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 integrity verifier — shed, queue, or backpressure — and make it explicit in code and in a test. Leaving the behavior to chance is how a point-in-time backup traffic spike becomes a data loss incident.

Testing

Each phase has a focused test: copy-on-write fork, chunk uploader, and manifest writer are exercised in isolation, then an integration test drives the full point-in-time backup pipeline and asserts the a chunk upload that case is handled.

Start with unit tests that pin each component’s contract: the copy-on-write fork returns the documented shape, the chunk uploader transforms correctly, and the manifest writer 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 point-in-time backup scenario end to end. Assert the happy path first, then deliberately trigger a chunk upload that fails partway and leaves the manifest referencing a missing block and assert the system degrades rather than corrupts. Because the components are pure where practical, the integration test stays deterministic rather than flaky.

A point-in-time backup 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 point-in-time backup 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 manifest writer failure-mode assertions live as tests rather than as a manual checklist that drifts out of date.

# run the point-in-time backup test suite
pnpm vitest run

Deployment and operations

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

The point-in-time backup system is a choke point, so it is the right place to reject malformed input before it reaches chunk uploader or manifest writer. 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 point-in-time backup 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 point-in-time backup system, and the lockfile is what makes the audit meaningful.

Observability

Emit one structured event per point-in-time backup 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 point-in-time backup decision and to the component log entry for the same operation. Without it, an investigation starts from a timestamp and a guess.

Watch the manifest writer failure rate separately from the overall error rate. A jump in manifest writer failures usually means a chunk upload that fails partway and leaves the manifest referencing a missing block, not a bug in chunk uploader, and the two have different owners and different fixes.

Chart the share of point-in-time backup 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 chunk upload that fails partway and leaves the manifest referencing a missing block. Reproduce with a minimal fixture, then add a guard test so the point-in-time backup system rejects the condition loudly.

Performance regresses under point-in-time backup load. Profile with the sampling analyzer, locate the hot path in chunk uploader, and add a targeted benchmark to lock the regression.

Keep the point-in-time backup dashboards separate from dependency dashboards. When an operation fails, the correlation id lets you decide in seconds whether the fault is copy-on-write fork, chunk uploader, or manifest writer, instead of guessing from a timeout.

Limitations and next steps

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

The point-in-time backup system is single-region by design; extending to multi-region means making copy-on-write fork shareable across regions, which is a real change to the consistency model rather than a config flag. The manifest writer 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 point-in-time backup correctness that is the point.

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