caching-strategy-optimizer（缓存策略优化器）

Caching Strategy Optimizer

Dramatically speed up CI pipelines with intelligent caching.

Cache Key Strategy

Package Manager Caches

# NPM - Hash package-lock.json
- uses: actions/cache@v3
  with:
    path: ~/.npm
    key: ${{ runner.os }}-npm-${{ hashFiles('**/package-lock.json') }}
    restore-keys: |
      ${{ runner.os }}-npm-

# pnpm - More aggressive caching
- uses: pnpm/action-setup@v2
  with:
    version: 8

- uses: actions/cache@v3
  with:
    path: |
      ~/.pnpm-store
      node_modules
    key: ${{ runner.os }}-pnpm-${{ hashFiles('**/pnpm-lock.yaml') }}
    restore-keys: |
      ${{ runner.os }}-pnpm-

# Python pip
- uses: actions/cache@v3
  with:
    path: ~/.cache/pip
    key: ${{ runner.os }}-pip-${{ hashFiles('**/requirements.txt') }}

# Cargo/Rust
- uses: actions/cache@v3
  with:
    path: |
      ~/.cargo/bin/
      ~/.cargo/registry/index/
      ~/.cargo/registry/cache/
      ~/.cargo/git/db/
      target/
    key: ${{ runner.os }}-cargo-${{ hashFiles('**/Cargo.lock') }}

Docker Layer Caching

Using Buildx

- name: Set up Docker Buildx
  uses: docker/setup-buildx-action@v3

- name: Build with cache
  uses: docker/build-push-action@v5
  with:
    context: .
    cache-from: type=gha
    cache-to: type=gha,mode=max

Registry-based Cache

- name: Build with registry cache
  uses: docker/build-push-action@v5
  with:
    context: .
    cache-from: type=registry,ref=myapp:buildcache
    cache-to: type=registry,ref=myapp:buildcache,mode=max

Build Output Caching

# Next.js cache
- uses: actions/cache@v3
  with:
    path: |
      ${{ github.workspace }}/.next/cache
    key: ${{ runner.os }}-nextjs-${{ hashFiles('**/package-lock.json') }}-${{ hashFiles('**/*.js', '**/*.jsx', '**/*.ts', '**/*.tsx') }}
    restore-keys: |
      ${{ runner.os }}-nextjs-${{ hashFiles('**/package-lock.json') }}-
      ${{ runner.os }}-nextjs-

# Webpack cache
- uses: actions/cache@v3
  with:
    path: node_modules/.cache/webpack
    key: ${{ runner.os }}-webpack-${{ hashFiles('webpack.config.js') }}-${{ hashFiles('src/**') }}

# TypeScript build cache
- uses: actions/cache@v3
  with:
    path: |
      dist
      tsconfig.tsbuildinfo
    key: ${{ runner.os }}-tsc-${{ hashFiles('**/*.ts') }}

Test Results Caching

# Jest cache
- uses: actions/cache@v3
  with:
    path: /tmp/jest_rt
    key: ${{ runner.os }}-jest-${{ hashFiles('**/*.test.ts') }}

# Pytest cache
- uses: actions/cache@v3
  with:
    path: .pytest_cache
    key: ${{ runner.os }}-pytest-${{ hashFiles('**/*test*.py') }}

Before/After Metrics

## Before Optimization

- Total time: 12 minutes
- npm ci: 4 minutes
- Build: 5 minutes
- Tests: 3 minutes

## After Caching

- Total time: 3 minutes
- npm ci: 30 seconds (cache hit)
- Build: 1 minute (incremental)
- Tests: 1.5 minutes (cache hit)

**Improvement: 75% faster (12m → 3m)**

Cache Hit Rate Monitoring

- name: Check cache hit
  id: cache
  uses: actions/cache@v3
  with:
    path: node_modules
    key: ${{ runner.os }}-deps-${{ hashFiles('package-lock.json') }}

- name: Log cache status
  run: |
    if [ "${{ steps.cache.outputs.cache-hit }}" == "true" ]; then
      echo "✅ Cache hit - saved $(date -u -d @$SECONDS +%M:%S)"
    else
      echo "❌ Cache miss - installing from scratch"
    fi

Best Practices

1. Precise keys: Include all dependencies in hash
2. Restore keys: Fallback to partial matches
3. Multiple paths: Cache related files together
4. Size limits: GitHub Actions limit is 10GB
5. Expiration: Caches expire after 7 days
6. Mode=max: Docker cache mode for better hits
7. Monitor hits: Track cache effectiveness

Common Pitfalls

❌ Too generic keys: key: deps (always hits) ✅ Specific keys: key: deps-${{hashFiles('package-lock.json')}}

❌ Missing restore-keys: Cache miss on minor changes ✅ Restore keys: Partial match fallback

❌ Caching node_modules with wrong lock file ✅ Match lock file: Hash the right lockfile

Output Checklist

• Package manager cache configured
• Build output cache
• Docker layer cache (if applicable)
• Test cache configured
• Cache keys use file hashes
• Restore keys for fallback
• Before/after metrics documented
• Cache hit monitoring