http-fetch Git Command Guide

The git http-fetch protocol implements client-side HTTP fetching for retrieving Git objects and references from remote repositories served over HTTP/HTTPS.

HTTP Fetch Protocol Features:

• Reference Discovery: Retrieve remote repository references via HTTP
• Efficient Transfer: Use smart HTTP protocol for optimized object fetching
• Authentication Support: Handle HTTP authentication and TLS certificates
• Incremental Updates: Support for shallow cloning and partial fetching
• Protocol Negotiation: Compatible with Git’s smart HTTPBackend servers

Protocol Components:

Discovery Phase:

GET /repo.git/info/refs?service=git-upload-pack  # Smart protocol
GET /repo.git/HEAD                              # Default branch
GET /repo.git/info/refs                         # Reference listing

Pack Transfer:

POST /repo.git/git-upload-pack                  # Upload pack negotiation
# Binary pack data transfer follows acknowledgment

Configuration Options:

Authentication Setup:

git config credential.helper store
git config http.sslVerify true
git config http.postBuffer 524288000  # 500MB for large transfers

Proxy Configuration:

git config http.proxy http://proxy.example.com:8080
git config https.proxy https://secure-proxy.example.com:8080

SSL/TLS Configuration:

git config http.sslCAInfo /path/to/ca-bundle.crt
git config http.sslCAPath /etc/ssl/certs/
git config http.sslCert /path/to/client-cert.pem

Troubleshooting Common Issues:

Authentication Problems:

# Check stored credentials
git config --list | grep credential

# Manual credential entry
git credential fill << EOF
protocol=https
host=github.com
EOF

SSL Certificate Issues:

# Disable SSL verification (not recommended)
git config http.sslVerify false

# Add custom CA certificate
git config http.sslCAInfo /path/to/ca-cert.pem

Caution

HTTP fetch requires proper server configuration. Smart HTTP protocol significantly more efficient than dumb HTTP. Authentication issues common with public repositories behind corporate firewalls.

git http-fetch Command Samples:

Basic repository clone via HTTP

git clone https://github.com/user/repo.git

Standard clone operation using HTTP fetch protocol.

Fetch updates from remote

git fetch origin

Retrieve latest changes using HTTP protocol.

Shallow clone for large repos

git clone --depth 1 https://github.com/user/repo.git

Limit history for faster initial fetch.

Clone with authentication

git clone https://username@github.com/user/repo.git

Fetch repository with username in URL.

Fetch specific branch

git fetch origin feature-branch:local-branch

Retrieve specific remote branch to local.

Update with custom buffer size

GIT_HTTP_BUFFER=524288000 git fetch origin

Increase buffer size for large repository fetches.

Fetch with proxy configuration

git -c http.proxy=http://proxy.example.com:8080 fetch origin

Fetch through HTTP proxy server.

Troubleshooting failed fetch

GIT_CURL_VERBOSE=1 git fetch origin 2>&1 | head -50

Enable verbose HTTP debugging output.

Tip

HTTP fetch works through firewalls that block SSH. Use credential helpers for secure authentication. Consider depth-limited clones for large repositories.

How does HTTP transfer differ from SSH?

HTTP uses regular web ports (80/443) while SSH uses port 22. HTTP can cache at network level but typically slower due to protocol overhead.

Why would HTTP fetch fail while SSH succeeds?

Common causes: corporate proxies blocking Git protocol, authentication token expiry, SSL certificate trust issues, or server-side rate limiting.

What’s smart HTTP protocol?

Optimized Git transfer over HTTP that uses Git’s native pack protocol instead of downloading individual files. Significantly faster than dumb HTTP.

How do I handle self-signed certificates?

Use git config http.sslCAInfo to specify custom CA certificate, or disable verification with http.sslVerify false (not recommended for production).

Can I resume interrupted HTTP fetches?

Partial pack file downloads cannot be resumed. Use git fetch —force to restart. For large repositories, consider shallow clones first.

What’s the impact of proxy on fetch performance?

Proxies add latency and may cache stale responses. HTTP caching works poorly for Git’s dynamic fetch requests. SSH typically faster through proxies.

How do I debug HTTP connection issues?

Set GIT_CURL_VERBOSE=1 for detailed connection logs, check firewall/proxy policies, verify SSL certificates, and test direct HTTP access to repository.

Applications of the http-fetch command

1. Firewall Traversal: Access repositories through corporate firewalls that block SSH
2. Public Repository Access: Standard method for cloning public Git repositories
3. CDN Compatibility: Leverage web infrastructure for distributed repository serving
4. Enterprise Environments: HTTP access through proxy servers and authenticated tunnels
5. Continuous Integration: Automated fetching in CI/CD pipelines behind corporate networks
6. Bandwidth Optimization: Smart HTTP’s efficient pack-based transfer for large repositories