Git daemon

Git daemon. It is a simple, stateless server program included with the Git (software) distributed version control system, designed specifically for providing anonymous, unauthenticated read-only access to repositories. Operating over the custom, efficient Git protocol, it is often used in tandem with other access methods like HTTP or SSH to serve public projects. Its lightweight and fast nature makes it a common choice for hosting large-scale open-source repositories, such as those on the Linux kernel or projects within the Apache Software Foundation.

Overview

The Git daemon was conceived as part of the original Git (software) toolkit to facilitate easy sharing of code without the overhead of user authentication. It implements the network side of the Git protocol, a simple data transfer mechanism optimized for the Delta encoding used internally by Git. Unlike more complex servers like Apache HTTP Server or services requiring Secure Shell, it has no built-in security layer, operating on the assumption that served data is intended for public consumption. This design philosophy aligns with the collaborative, open-source ethos of projects like those hosted by the Free Software Foundation and mirrors the early distribution methods of systems like BitKeeper.

Operation

By default, the daemon listens on port 9418, as assigned by the Internet Assigned Numbers Authority. When started, typically via the `git daemon` command, it can be configured to export all repositories within a specified directory tree. A client, such as another Git (software) installation, can then initiate a fetch operation using a URL beginning with `git://`. The communication involves the server advertising references and the client requesting Packfile objects. For large transfers, such as cloning the Android Open Source Project, the efficiency of this protocol can significantly outperform HTTP-based methods. The daemon's process is managed by inetd or systemd on many Linux distributions, and it can also run as a standalone service on Microsoft Windows via Cygwin or Windows Subsystem for Linux.

Configuration

Configuration is primarily handled through command-line options and the presence of special files within the repository directory. The `--export-all` option instructs the daemon to serve any repository, while the `--base-path` defines the root directory for exported paths. To restrict access, an administrator can create a file named `git-daemon-export-ok` inside a repository; only repositories containing this marker file will be served if `--export-all` is not used. Fine-grained control over enabled services, such as `upload-pack` for fetching, is also possible. These settings are often integrated into larger infrastructure automation tools used by organizations like GitHub or GitLab for their internal mirroring networks.

Security considerations

Due to its lack of authentication and encryption, the Git daemon is unsuitable for serving private or sensitive data. The protocol transmits all data, including commit histories and Source code, in plaintext, making it vulnerable to Packet sniffing on untrusted networks. Consequently, it is almost always deployed behind a Firewall (computing) on a trusted segment, or used in conjunction with a Reverse proxy like Nginx for access control and Load balancing (computing). For secure operations, the Git (software) ecosystem strongly recommends using the SSH protocol or HTTPS with authentication. Major hosting providers like Microsoft with Azure DevOps and Atlassian with Bitbucket do not offer the native Git protocol for this reason, defaulting to more secure channels.

Use cases

The primary use case is providing high-performance, public read access to large open-source repositories. The Linux kernel project and its mirrors historically relied heavily on this daemon for distribution. It is also ideal for creating internal network mirrors or caches within an organization to reduce bandwidth usage and improve clone speeds for developers; tools like Gitolite can manage authorization while the daemon handles efficient data transfer. Furthermore, it serves as the backbone for some Peer-to-peer Git hosting experiments and is used in automated testing environments, such as those built on Jenkins (software) or Travis CI, where fast, anonymous repository access is required for building and deploying software.

Category:Git (software) Category:Server software Category:Free software programmed in C