git-pack-redundant

git pack-redundant [pack-file...]

pack-file...
    One or more paths to Git packfiles (e.g., .git/objects/pack/pack-abcdef.pack) to be checked for redundancy. If this argument is omitted, git-pack-redundant will examine all packfiles present in the repository's object database (typically located in the .git/objects/pack/ directory) to find those that are redundant relative to each other.

The git-pack-redundant command is a Git plumbing utility designed to find and report packfiles whose objects are entirely contained within other packfiles. A packfile is considered redundant if every single object it contains is also present in one or more other packfiles specified in the input, or within the repository's entire object database if no specific packfiles are listed.

This command is vital for the efficient maintenance and cleanup of a Git repository's object database. It helps in identifying and marking obsolete or duplicate packfiles that may accumulate over time due to operations like fetches, pushes, or repository repacks. For instance, commands such as git repack -d leverage git-pack-redundant internally to safely remove old packfiles after a new, optimized pack has been created and verified. The output consists of a list of filenames of the identified redundant packfiles, one per line.

git-pack-redundant strictly identifies packfiles that are entirely redundant. If a packfile contains even a single object that is not present in any other packfile under consideration, it will not be reported as redundant.

Due to the nature of its operation, this command can be computationally intensive for repositories with a very large number of packfiles, as it needs to open and analyze their respective index files (.idx) to determine object containment. It is primarily an internal Git command, not typically run directly by end-users.

The command outputs the full paths and filenames of the identified redundant packfiles, with each filename on a new line, to standard output. This output can be piped to other commands for automated deletion or processing.

While runnable manually, git-pack-redundant's primary role is within Git's automated maintenance scripts and other commands. For example, after running git repack -ad, the command is used to safely identify and remove the old packfiles whose objects have been moved into the newly created, optimized pack.

As a fundamental plumbing command, git-pack-redundant has been an integral part of Git's object database management since the early stages of its development. Its purpose is to support robust and efficient repository maintenance, particularly in the automated cleanup of obsolete packfiles that result from operations consolidating or optimizing the object store. Its focus remains on providing a precise mechanism for identifying completely covered packfiles, contributing to the overall compactness and performance of Git repositories.

git-repack(1), git-prune-packed(1), git-verify-pack(1), git-pack-objects(1)