configure-git-webserver

Git Web Server Deployment Configuration

This skill provides guidance for configuring Git-based web deployment systems where content is automatically deployed to a web server when pushed to a Git repository.

Task Pattern Recognition

This skill applies to tasks involving:

• Setting up bare Git repositories for deployment
• Configuring post-receive hooks for automated deployment
• Running web servers to serve deployed content
• Creating push-to-deploy workflows (e.g., git push triggers deployment)
• SSH-based remote Git access configuration

Pre-Implementation Checklist

Before starting implementation, verify the environment:

1. Check available tools and permissions

   • Determine if sudo is available: command -v sudo
   • Check available web servers: command -v nginx, command -v python3
   • Verify git is installed: git --version
   • Check current user context: whoami, id

2. Configure Git identity proactively

   • Set user.name and user.email before any Git operations
   • Prevents errors during testing phase: git config --global user.name "Deploy" && git config --global user.email "deploy@localhost"

3. Identify port availability

   • Check if intended port is in use: lsof -i :PORT or ss -tlnp | grep PORT
   • Have fallback ports ready (8080, 8000, 3000)

Implementation Approach

Step 1: Create Bare Git Repository

# Create the bare repository directory
mkdir -p /git/server
cd /git/server
git init --bare

A bare repository has no working directory and is designed for remote operations.

Step 2: Create Web Directory

# Create web root with appropriate permissions
mkdir -p /var/www/html
chmod 755 /var/www/html

Document why this directory was chosen (standard location, permissions, etc.).

Step 3: Create Post-Receive Hook

The post-receive hook triggers after each push. Critical considerations:

#!/bin/bash
# /git/server/hooks/post-receive

# Handle multiple branch naming conventions
while read oldrev newrev refname; do
    branch=$(echo $refname | sed 's/refs\/heads\///')

    if [ "$branch" = "main" ] || [ "$branch" = "master" ]; then
        GIT_WORK_TREE=/var/www/html git checkout -f $branch
        echo "Deployed $branch to /var/www/html"
    fi
done

Make executable: chmod +x /git/server/hooks/post-receive

Hook considerations:

• Handle both main and master branch names
• Use force checkout (-f) to overwrite local changes
• Log deployment actions for debugging
• Consider adding error handling and notifications

Step 4: Web Server Setup

Option A: Python Simple Server (Development/Testing)

cd /var/www/html
python3 -m http.server 8080 &

Limitations: No auto-restart, no production features, terminates with session.

Option B: Production-Ready Setup

For persistent service, create a systemd service or use a process manager. See references/deployment_checklist.md for service configuration patterns.

Step 5: Verification Strategy

Test the actual use case, not just components:

1. Local clone test (basic validation):

   git clone /git/server /tmp/test-clone
   cd /tmp/test-clone
   echo "test" > index.html
   git add . && git commit -m "test" && git push origin main
   

2. Remote SSH test (full validation):

   git clone user@server:/git/server /tmp/remote-test
   

   This validates the SSH access path that users will actually use.

3. Web server verification:

   curl http://localhost:8080/index.html
   

4. Cleanup test artifacts:

   rm -rf /tmp/test-clone /tmp/remote-test
   

Common Pitfalls and Solutions

1. Service Persistence

Problem: Background processes (command &) terminate when session ends.

Solution:

• Use systemd services, supervisord, or similar process managers
• At minimum, document that the server requires manual restart
• Consider using nohup for slightly better persistence: nohup python3 -m http.server 8080 &

2. Permission Issues

Problem: Different users pushing may not have write access to web directory.

Solutions:

• Ensure web directory is writable by git user
• Consider group permissions: chgrp -R webgroup /var/www/html && chmod -R g+w /var/www/html
• Verify SSH user matches repository permissions

3. Missing Git Configuration

Problem: git commit fails during testing without user.name/user.email.

Solution: Configure Git identity at the start of implementation, not when errors occur.

4. Port Conflicts

Problem: Web server fails to start because port is already in use.

Solution:

• Check port availability before starting server
• Use alternative ports if needed
• Kill existing processes if appropriate: fuser -k 8080/tcp

5. Testing Wrong Workflow

Problem: Testing with local paths doesn't validate SSH-based remote access.

Solution: Test both local (/git/server) and remote (user@server:/git/server) access patterns when SSH is part of the requirements.

6. Hook Not Executing

Problem: Post-receive hook doesn't run after push.

Checklist:

• Hook is executable: chmod +x hooks/post-receive
• Hook has correct shebang: #!/bin/bash
• No syntax errors: bash -n hooks/post-receive
• Check hook output in push response

7. Force Checkout Overwrites

Problem: git checkout -f overwrites files modified directly in web directory.

Consideration: This is usually desired behavior for deployment, but document it for users who may manually edit deployed files.

Verification Checklist

Before marking task complete, verify:

• Bare repository initialized and accessible
• Post-receive hook is executable and tested
• Web directory exists with correct permissions
• Web server is running and serving content
• End-to-end push-to-deploy workflow tested
• Service persistence documented or configured
• Test artifacts cleaned up
• Remote access path validated (if SSH-based)

Environment-Specific Considerations

• Container environments: Services may need different persistence strategies
• Restricted environments: sudo may not be available; adapt paths accordingly
• Multi-user systems: Consider shared access and permission inheritance