How to fix error: RPC failed; HTTP 503 Service Unavailable in Git

GitBEGINNERMEDIUMUpdated Dec 6, 2025· 196 views

The 'RPC failed; HTTP 503 Service Unavailable' error occurs when Git cannot complete an operation because the remote server (GitHub, GitLab, Bitbucket, etc.) is temporarily unavailable or overloaded. This is typically a server-side issue that resolves itself, but there are steps you can take to work around it.

Quick fix

Fixes error: RPC failed; HTTP 503 Service Unavailable

# Wait 2-5 minutes, then retry your command
git push origin main

# Or for fetch/pull
git fetch origin
  1. Check the service status page
  2. Wait and retry the operation
  3. Try using SSH instead of HTTPS
  4. Check for rate limiting
git — error
# Wait 2-5 minutes, then retry your command
git push origin main
# Or for fetch/pull
git fetch origin
error: RPC failed; HTTP 503 Service Unavailable

What this error means

The "RPC failed; HTTP 503 Service Unavailable" error indicates that the remote Git server returned an HTTP 503 status code during a Git operation (clone, push, pull, or fetch). RPC (Remote Procedure Call) is the mechanism Git uses to communicate with remote servers over HTTP/HTTPS. HTTP 503 is a server-side error code meaning "Service Unavailable." This typically occurs when: - The server is temporarily overloaded with requests - The server is undergoing maintenance - A backend service the Git server depends on is down - Rate limiting has been applied to your requests - The server is experiencing technical difficulties Unlike client-side errors (4xx codes), a 503 error is not caused by anything wrong with your local Git configuration or credentials. The server is explicitly telling Git that it cannot handle the request at this time.

Typical symptoms

  • git push, pull, fetch, or clone fails with 'HTTP 503 Service Unavailable'
  • Error appears suddenly after Git operations were previously working
  • Multiple team members experience the same error simultaneously
  • Error occurs intermittently - sometimes it works, sometimes it fails
  • Operations on different repositories fail at the same time
  • The error often resolves itself after waiting a few minutes
  • Git operations via SSH may still work while HTTPS fails

Common causes

  1. Remote Git hosting service (GitHub, GitLab, Bitbucket) is experiencing an outage
  2. Server is undergoing scheduled or emergency maintenance
  3. High traffic or load on the Git server causing temporary unavailability
  4. Rate limiting applied due to too many requests from your IP or account
  5. Network infrastructure issues between you and the Git server
  6. Self-hosted Git server has crashed or is overloaded
  7. CDN or load balancer issues in front of the Git service
  8. Backend database or storage service used by the Git server is unavailable

How to fix "error: RPC failed; HTTP 503 Service Unavailable"

1Check the service status page

First, verify if the Git hosting service is experiencing issues:

GitHub:
- Status page: https://www.githubstatus.com/
- Twitter: https://twitter.com/githubstatus

GitLab:
- Status page: https://status.gitlab.com/

Bitbucket:
- Status page: https://bitbucket.status.atlassian.com/

Azure DevOps:
- Status page: https://status.dev.azure.com/

If the status page shows an incident, wait for the service to be restored. You can subscribe to status updates to be notified when the issue is resolved.

2Wait and retry the operation

HTTP 503 errors are usually temporary. Wait a few minutes and try again:

bash
# Wait 2-5 minutes, then retry your command
git push origin main

# Or for fetch/pull
git fetch origin

If you're running automated scripts or CI/CD pipelines, implement retry logic with exponential backoff:

bash
# Simple bash retry loop
for i in {1..5}; do
  git push origin main && break
  echo "Attempt $i failed, retrying in $((i * 30)) seconds..."
  sleep $((i * 30))
done
3Try using SSH instead of HTTPS

If HTTPS is failing due to server issues, SSH connections might still work as they often use different infrastructure:

bash
# Change remote URL from HTTPS to SSH
git remote set-url origin git@github.com:username/repo.git

# Verify the change
git remote -v

# Now try your operation
git push origin main

For GitLab:

bash
git remote set-url origin git@gitlab.com:username/repo.git

For Bitbucket:

bash
git remote set-url origin git@bitbucket.org:username/repo.git

Note: You'll need SSH keys set up for this to work. See the Git documentation for SSH key setup.

4Check for rate limiting

Some Git services return 503 errors when you've hit rate limits. Check if you're being rate limited:

For GitHub:

bash
# Check your rate limit status (requires authentication)
curl -H "Authorization: token YOUR_TOKEN" https://api.github.com/rate_limit

If you're hitting rate limits:
- Reduce the frequency of automated operations
- Use authenticated requests (they have higher limits)
- Implement caching for frequently accessed data
- Consider using GitHub Apps for higher rate limits

For CI/CD systems:
- Ensure you're using service accounts with appropriate rate limits
- Cache dependencies and repositories between builds
- Use shallow clones to reduce data transfer

5Reduce repository size for large operations

If you're cloning or fetching large repositories, the server might time out. Try reducing the data transfer:

bash
# Use shallow clone with limited history
git clone --depth 1 https://github.com/username/repo.git

# For existing repos, fetch with depth limit
git fetch --depth 1 origin main

# Use single branch clone
git clone --single-branch --branch main https://github.com/username/repo.git

For pushing large changes:

bash
# Push in smaller batches
git push origin main~10:main
git push origin main~5:main
git push origin main
6Check your network connection and proxy settings

Network issues between you and the server can sometimes manifest as 503 errors:

bash
# Test basic connectivity
ping github.com
curl -I https://github.com

# Check if a proxy is configured that might be failing
git config --global --get http.proxy
git config --global --get https.proxy

# Temporarily unset proxy to test direct connection
git config --global --unset http.proxy
git config --global --unset https.proxy

If you're behind a corporate proxy or firewall, it might be blocking or rate-limiting Git traffic. Contact your network administrator.

7For self-hosted servers: Check server health

If you're using a self-hosted Git server (GitLab, Gitea, Gogs, etc.):

bash
# Check server logs
sudo journalctl -u gitlab-runsvdir -f  # For GitLab
sudo journalctl -u gitea -f             # For Gitea

# Check system resources
top -b -n 1 | head -20
df -h  # Check disk space
free -m  # Check memory

# Restart the Git service if needed
sudo gitlab-ctl restart  # For GitLab
sudo systemctl restart gitea  # For Gitea

Common issues with self-hosted servers:
- Disk space full (especially for Git objects)
- Memory exhaustion
- Database connection issues
- Too many concurrent connections

8Use a Git mirror or alternative remote

If one remote is unavailable, you might be able to push to a mirror:

bash
# Add a secondary remote as backup
git remote add backup git@gitlab.com:username/repo.git

# Push to backup when primary is down
git push backup main

# Later, sync the mirrors when primary is back up
git push origin main

For read operations, you can also clone from mirrors:

bash
# Many popular repositories have mirrors
# Check if a mirror exists for the repository you need

Advanced notes

### Understanding 503 Errors in Distributed Systems

A 503 error from a Git server typically indicates the server is intentionally rejecting requests to protect itself. This is often a circuit breaker pattern in action - when a service becomes overloaded, it temporarily rejects new requests to prevent complete failure.

### GitHub-Specific Behavior

GitHub may return 503 errors during:
- High-traffic events (major open-source releases, popular projects)
- DDoS mitigation
- Deployment of new infrastructure
- Database maintenance windows

GitHub's smart HTTP protocol can sometimes trigger 503s during large operations even when the web interface works fine.

### CI/CD Pipeline Considerations

For CI/CD systems experiencing 503 errors:

yaml
# GitLab CI example with retry
job:
  script:
    - git push origin main
  retry:
    max: 3
    when:
      - runner_system_failure
      - stuck_or_timeout_failure

# GitHub Actions example
- name: Push changes
  run: |
    for i in 1 2 3 4 5; do
      git push origin main && break
      sleep $((i * 60))
    done

### HTTP Retry-After Header

Some servers include a Retry-After header in 503 responses indicating when to retry. Check verbose output:

bash
GIT_CURL_VERBOSE=1 git push origin main 2>&1 | grep -i retry

### Graceful Degradation Strategies

For production systems:
1. Implement circuit breakers in your Git automation
2. Set up repository mirroring to multiple providers
3. Cache Git objects locally using Git's partial clone feature
4. Use Git bundles for offline backup

bash
# Create a bundle backup
git bundle create repo-backup.bundle --all

# Restore from bundle if needed
git clone repo-backup.bundle restored-repo

Frequently asked questions

What does "error: RPC failed; HTTP 503 Service Unavailable" mean?
The "RPC failed; HTTP 503 Service Unavailable" error indicates that the remote Git server returned an HTTP 503 status code during a Git operation (clone, push, pull, or fetch). RPC (Remote Procedure Call) is the mechanism Git uses to communicate with remote servers over HTTP/HTTPS. " This typically occurs when: - The server is temporarily overloaded with requests - The server is undergoing maintenance - A backend service the Git server depends on is down - Rate limiting has been applied to your requests - The server is experiencing technical difficulties Unlike client-side errors (4xx codes), a 503 error is not caused by anything wrong with your local Git configuration or credentials.
How do I fix "error: RPC failed; HTTP 503 Service Unavailable"?
The 'RPC failed; HTTP 503 Service Unavailable' error occurs when Git cannot complete an operation because the remote server (GitHub, GitLab, Bitbucket, etc.) is temporarily unavailable or overloaded. This is typically a server-side issue that resolves itself, but there are steps you can take to work around it.
What causes "error: RPC failed; HTTP 503 Service Unavailable"?
Remote Git hosting service (GitHub, GitLab, Bitbucket) is experiencing an outage Server is undergoing scheduled or emergency maintenance High traffic or load on the Git server causing temporary unavailability
Is the "error: RPC failed; HTTP 503 Service Unavailable" error serious?
It is a medium-severity error — more of a blocker than a risk. The steps above resolve it.

Official resources & further reading