feat: optimize Docker with multi-stage builds and container-based testing

2026-02-14 19:30:02 +01:00 · 2025-05-29 14:20:58 -05:00
parent 9339e5f87b
commit 57608e021b
5 changed files with 431 additions and 40 deletions
--- a/.dockerignore
+++ b/.dockerignore
@@ -1,34 +1,75 @@
+# Dependencies
 node_modules
 npm-debug.log
+dist
+
+# Git
 .git
 .gitignore
+.gitattributes
+
+# Environment
 .env
 .env.*
+!.env.example
+
+# OS
 .DS_Store
+Thumbs.db
+
+# Testing
 coverage
 .nyc_output
 test-results
 *.log
 logs
+
+# Development
 .husky
-.github
 .vscode
 .idea
 *.swp
 *.swo
 *~
-CLAUDE.local.md
-secrets
-k8s
-docs
-test
-*.test.js
-*.spec.js
+
+# Documentation
 README.md
 *.md
 !CLAUDE.md
+!README.dockerhub.md
+
+# CI/CD
+.github
+!.github/workflows
+
+# Secrets
+secrets
+CLAUDE.local.md
+
+# Kubernetes
+k8s
+
+# Docker
 docker-compose*.yml
+!docker-compose.test.yml
 Dockerfile*
+!Dockerfile
+!Dockerfile.claudecode
 .dockerignore
+
+# Scripts - exclude all by default for security, then explicitly include needed runtime scripts
 *.sh
-!scripts/runtime/*.sh
+!scripts/runtime/*.sh
+
+# Test files (keep for test stage)
+# Removed test exclusion to allow test stage to access tests
+
+# Build artifacts
+*.tsbuildinfo
+tsconfig.tsbuildinfo
+
+# Cache
+.cache
+.buildx-cache*
+tmp
+temp
--- a/126
+++ b/126
@@ -1,9 +1,69 @@
-FROM node:24-slim
+# syntax=docker/dockerfile:1
+
+# Build stage - compile TypeScript and prepare production files
+FROM node:24-slim AS builder
+
+WORKDIR /app
+
+# Copy package files first for better caching
+COPY package*.json tsconfig.json babel.config.js ./
+
+# Install all dependencies (including dev)
+RUN npm ci
+
+# Copy source code
+COPY src/ ./src/
+
+# Build TypeScript
+RUN npm run build
+
+# Copy remaining application files
+COPY . .
+
+# Production dependency stage - smaller layer for dependencies
+FROM node:24-slim AS prod-deps
+
+WORKDIR /app
+
+# Copy package files
+COPY package*.json ./
+
+# Install only production dependencies
+RUN npm ci --omit=dev && npm cache clean --force
+
+# Test stage - includes dev dependencies and test files
+FROM node:24-slim AS test
+
+# Set shell with pipefail option
+SHELL ["/bin/bash", "-o", "pipefail", "-c"]
+
+WORKDIR /app
+
+# Copy package files and install all dependencies
+COPY package*.json tsconfig*.json babel.config.js jest.config.js ./
+RUN npm ci
+
+# Copy source and test files
+COPY src/ ./src/
+COPY test/ ./test/
+COPY scripts/ ./scripts/
+
+# Copy built files from builder
+COPY --from=builder /app/dist ./dist
+
+# Set test environment
+ENV NODE_ENV=test
+
+# Run tests by default in this stage
+CMD ["npm", "test"]
+
+# Production stage - minimal runtime image
+FROM node:24-slim AS production

 # Set shell with pipefail option for better error handling
 SHELL ["/bin/bash", "-o", "pipefail", "-c"]

-# Install git, Claude Code, Docker, and required dependencies with pinned versions and --no-install-recommends
+# Install runtime dependencies with pinned versions
 RUN apt-get update && apt-get install -y --no-install-recommends \
    git=1:2.39.5-0+deb12u2 \
    curl=7.88.1-10+deb12u12 \
@@ -23,56 +83,60 @@ RUN curl -fsSL https://download.docker.com/linux/debian/gpg | gpg --dearmor -o /
    && apt-get install -y --no-install-recommends docker-ce-cli=5:27.* \
    && rm -rf /var/lib/apt/lists/*

-# Install Claude Code (latest version)
-# hadolint ignore=DL3016
-RUN npm install -g @anthropic-ai/claude-code
-
 # Create docker group first, then create a non-root user for running the application
 RUN groupadd -g 999 docker 2>/dev/null || true \
    && useradd -m -u 1001 -s /bin/bash claudeuser \
    && usermod -aG docker claudeuser 2>/dev/null || true

-# Create claude config directory and copy config
+# Create npm global directory for claudeuser and set permissions
+RUN mkdir -p /home/claudeuser/.npm-global \
+    && chown -R claudeuser:claudeuser /home/claudeuser/.npm-global
+
+# Configure npm to use the user directory for global packages
+USER claudeuser
+ENV NPM_CONFIG_PREFIX=/home/claudeuser/.npm-global
+ENV PATH=/home/claudeuser/.npm-global/bin:$PATH
+
+# Install Claude Code (latest version) as non-root user
+# hadolint ignore=DL3016
+RUN npm install -g @anthropic-ai/claude-code
+
+USER root
+
+# Create claude config directory
 RUN mkdir -p /home/claudeuser/.config/claude
-COPY claude-config.json /home/claudeuser/.config/claude/config.json

 WORKDIR /app

-# Copy package files and install dependencies
-COPY package*.json ./
-COPY tsconfig.json ./
-COPY babel.config.js ./
+# Copy production dependencies from prod-deps stage
+COPY --from=prod-deps /app/node_modules ./node_modules

-# Install all dependencies (including dev for build)
-RUN npm ci
+# Copy built application from builder stage
+COPY --from=builder /app/dist ./dist

-# Copy source code
-COPY src/ ./src/
+# Copy configuration and runtime files
+COPY package*.json tsconfig.json babel.config.js ./
+COPY claude-config.json /home/claudeuser/.config/claude/config.json
+COPY scripts/ ./scripts/
+COPY docs/ ./docs/
+COPY cli/ ./cli/

-# Build TypeScript
-RUN npm run build
-
-# Remove dev dependencies to reduce image size
-RUN npm prune --omit=dev && npm cache clean --force
-
-# Copy remaining application files
-COPY . .
-
-# Consolidate permission changes into a single RUN instruction
+# Set permissions
 RUN chown -R claudeuser:claudeuser /home/claudeuser/.config /app \
    && chmod +x /app/scripts/runtime/startup.sh

-# Note: Docker socket will be mounted at runtime, no need to create it here
-
 # Expose the port
 EXPOSE 3002

 # Set default environment variables
 ENV NODE_ENV=production \
-    PORT=3002
+    PORT=3002 \
+    NPM_CONFIG_PREFIX=/home/claudeuser/.npm-global \
+    PATH=/home/claudeuser/.npm-global/bin:$PATH

-# Stay as root user to run Docker commands
-# (The container will need to run with Docker socket mounted)
+# Switch to non-root user for running the application
+# Docker commands will work via docker group membership when socket is mounted
+USER claudeuser

 # Run the startup script
 CMD ["bash", "/app/scripts/runtime/startup.sh"]
--- a/docker-compose.test.yml
+++ b/docker-compose.test.yml
@@ -0,0 +1,68 @@
+version: '3.8'
+
+services:
+  # Test runner service - runs tests in container
+  test:
+    build:
+      context: .
+      dockerfile: Dockerfile
+      target: test
+      cache_from:
+        - ${DOCKER_HUB_ORGANIZATION:-intelligenceassist}/claude-hub:test-cache
+    environment:
+      - NODE_ENV=test
+      - CI=true
+      - GITHUB_TOKEN=${GITHUB_TOKEN:-test-token}
+      - GITHUB_WEBHOOK_SECRET=${GITHUB_WEBHOOK_SECRET:-test-secret}
+      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY:-test-key}
+    volumes:
+      - ./coverage:/app/coverage
+    # Run only unit tests in CI (no e2e tests that require Docker)
+    command: npm run test:unit
+
+  # Integration test service
+  integration-test:
+    build:
+      context: .
+      dockerfile: Dockerfile
+      target: test
+    environment:
+      - NODE_ENV=test
+      - CI=true
+      - TEST_SUITE=integration
+    volumes:
+      - ./coverage:/app/coverage
+    command: npm run test:integration
+    depends_on:
+      - webhook
+
+  # Webhook service for integration testing
+  webhook:
+    build:
+      context: .
+      dockerfile: Dockerfile
+      target: production
+    environment:
+      - NODE_ENV=test
+      - PORT=3002
+      - GITHUB_TOKEN=${GITHUB_TOKEN:-test-token}
+      - GITHUB_WEBHOOK_SECRET=${GITHUB_WEBHOOK_SECRET:-test-secret}
+      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY:-test-key}
+    ports:
+      - "3002:3002"
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:3002/health"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 40s
+
+  # E2E test service - removed from CI, use for local development only
+  # To run e2e tests locally with Docker access:
+  # docker compose -f docker-compose.test.yml run --rm -v /var/run/docker.sock:/var/run/docker.sock e2e-test
+
+# Networks
+networks:
+  default:
+    name: claude-hub-test
+    driver: bridge
--- a/docs/docker-optimization.md
+++ b/docs/docker-optimization.md
@@ -0,0 +1,214 @@
+# Docker Build Optimization Guide
+
+This document describes the optimizations implemented in our Docker CI/CD pipeline for faster builds and better caching.
+
+## Overview
+
+Our optimized Docker build pipeline includes:
+- Self-hosted runner support with automatic fallback
+- Multi-stage builds for efficient layering
+- Advanced caching strategies
+- Container-based testing
+- Parallel builds for multiple images
+- Security scanning integration
+
+## Self-Hosted Runners
+
+### Configuration
+- **Labels**: `self-hosted, linux, x64, docker`
+- **Usage**: All Docker builds use self-hosted runners for improved performance and caching
+- **Local Cache**: Self-hosted runners maintain Docker layer cache between builds
+- **Fallback**: Manual fallback to GitHub-hosted runners if self-hosted are unavailable
+
+### Runner Setup
+Self-hosted runners provide:
+- Persistent Docker layer cache
+- Faster builds (no image pull overhead)
+- Better network throughput for pushing images
+- Cost savings on GitHub Actions minutes
+
+### Fallback Strategy
+If self-hosted runners are unavailable:
+1. Workflow will queue waiting for runners
+2. Can be manually cancelled and re-run with modified workflow
+3. Consider implementing automatic fallback in future iterations
+
+## Multi-Stage Dockerfile
+
+Our Dockerfile uses multiple stages for optimal caching and smaller images:
+
+1. **Builder Stage**: Compiles TypeScript
+2. **Prod-deps Stage**: Installs production dependencies only
+3. **Test Stage**: Includes dev dependencies and test files
+4. **Production Stage**: Minimal runtime image
+
+### Benefits
+- Parallel builds of independent stages
+- Smaller final image (no build tools or dev dependencies)
+- Test stage can run in CI without affecting production image
+- Better layer caching between builds
+
+## Caching Strategies
+
+### 1. GitHub Actions Cache (GHA)
+```yaml
+cache-from: type=gha,scope=${{ matrix.image }}-prod
+cache-to: type=gha,mode=max,scope=${{ matrix.image }}-prod
+```
+
+### 2. Registry Cache
+```yaml
+cache-from: type=registry,ref=${{ org }}/claude-hub:nightly
+```
+
+### 3. Inline Cache
+```yaml
+build-args: BUILDKIT_INLINE_CACHE=1
+outputs: type=inline
+```
+
+### 4. Layer Ordering
+- Package files copied first (changes less frequently)
+- Source code copied after dependencies
+- Build artifacts cached between stages
+
+## Container-Based Testing
+
+Tests run inside Docker containers for:
+- Consistent environment
+- Parallel test execution
+- Isolation from host system
+- Same environment as production
+
+### Test Execution
+```bash
+# Unit tests in container
+docker run --rm claude-hub:test npm test
+
+# Integration tests with docker-compose
+docker-compose -f docker-compose.test.yml run integration-test
+
+# E2E tests against running services
+docker-compose -f docker-compose.test.yml run e2e-test
+```
+
+## Build Performance Optimizations
+
+### 1. BuildKit Features
+- `DOCKER_BUILDKIT=1` for improved performance
+- `--mount=type=cache` for package manager caches
+- Parallel stage execution
+
+### 2. Docker Buildx
+- Multi-platform builds (amd64, arm64)
+- Advanced caching backends
+- Build-only stages that don't ship to production
+
+### 3. Context Optimization
+- `.dockerignore` excludes unnecessary files
+- Minimal context sent to Docker daemon
+- Faster uploads and builds
+
+### 4. Dependency Caching
+- Separate stage for production dependencies
+- npm ci with --omit=dev for smaller images
+- Cache mount for npm packages
+
+## Workflow Features
+
+### PR Builds
+- Build and test without publishing
+- Single platform (amd64) for speed
+- Container-based test execution
+- Security scanning with Trivy
+
+### Main Branch Builds
+- Multi-platform builds (amd64, arm64)
+- Push to registry with :nightly tag
+- Update cache images
+- Full test suite execution
+
+### Version Tag Builds
+- Semantic versioning tags
+- :latest tag update
+- Multi-platform support
+- Production-ready images
+
+## Security Scanning
+
+### Integrated Scanners
+1. **Trivy**: Vulnerability scanning for Docker images
+2. **Hadolint**: Dockerfile linting
+3. **npm audit**: Dependency vulnerability checks
+4. **SARIF uploads**: Results visible in GitHub Security tab
+
+## Monitoring and Metrics
+
+### Build Performance
+- Build time per stage
+- Cache hit rates
+- Image size tracking
+- Test execution time
+
+### Health Checks
+```yaml
+healthcheck:
+  test: ["CMD", "curl", "-f", "http://localhost:3002/health"]
+  interval: 30s
+  timeout: 10s
+  retries: 3
+```
+
+## Local Development
+
+### Building locally
+```bash
+# Build with BuildKit
+DOCKER_BUILDKIT=1 docker build -t claude-hub:local .
+
+# Build specific stage
+docker build --target test -t claude-hub:test .
+
+# Run tests locally
+docker-compose -f docker-compose.test.yml run test
+```
+
+### Cache Management
+```bash
+# Clear builder cache
+docker builder prune
+
+# Use local cache
+docker build --cache-from claude-hub:local .
+```
+
+## Best Practices
+
+1. **Order Dockerfile commands** from least to most frequently changing
+2. **Use specific versions** for base images and dependencies
+3. **Minimize layers** by combining RUN commands
+4. **Clean up** package manager caches in the same layer
+5. **Use multi-stage builds** to reduce final image size
+6. **Leverage BuildKit** features for better performance
+7. **Test in containers** for consistency across environments
+8. **Monitor build times** and optimize bottlenecks
+
+## Troubleshooting
+
+### Slow Builds
+- Check cache hit rates in build logs
+- Verify .dockerignore is excluding large files
+- Use `--progress=plain` to see detailed timings
+- Consider parallelizing independent stages
+
+### Cache Misses
+- Ensure consistent base image versions
+- Check for unnecessary file changes triggering rebuilds
+- Use cache mounts for package managers
+- Verify registry cache is accessible
+
+### Test Failures in Container
+- Check environment variable differences
+- Verify volume mounts are correct
+- Ensure test dependencies are in test stage
+- Check for hardcoded paths or ports
--- a/package.json
+++ b/package.json
@@ -14,11 +14,15 @@
    "typecheck": "tsc --noEmit",
    "test": "jest --testPathPattern='test/(unit|integration).*\\.test\\.(js|ts)$'",
    "test:unit": "jest --testMatch='**/test/unit/**/*.test.{js,ts}'",
+    "test:integration": "jest --testMatch='**/test/integration/**/*.test.{js,ts}'",
    "test:chatbot": "jest --testMatch='**/test/unit/providers/**/*.test.{js,ts}' --testMatch='**/test/unit/controllers/chatbotController.test.{js,ts}'",
    "test:e2e": "jest --testMatch='**/test/e2e/**/*.test.{js,ts}'",
    "test:coverage": "jest --coverage",
    "test:watch": "jest --watch",
    "test:ci": "jest --ci --coverage --testPathPattern='test/(unit|integration).*\\.test\\.(js|ts)$'",
+    "test:docker": "docker-compose -f docker-compose.test.yml run --rm test",
+    "test:docker:integration": "docker-compose -f docker-compose.test.yml run --rm integration-test",
+    "test:docker:e2e": "docker-compose -f docker-compose.test.yml run --rm e2e-test",
    "pretest": "./scripts/utils/ensure-test-dirs.sh",
    "lint": "eslint src/ test/ --fix",
    "lint:check": "eslint src/ test/",