Maintaining a Git Repository
Maintenance of your Git repository typically involves reducing a repository's size. If you imported from another version control system, you may need to clean up unnecessary files after the import. This page focuses on removing large files from a Git repo and contains the following topics:
Be very careful...
The procedure and tools on this page use advanced techniques that involve destructive operations. Make sure you read carefully and backup your repository before starting. The easiest way to create a backup is to clone your repository using the --mirror flag, and zip the whole clone. With the backup, if you accidentally corrupt a key element of your repo during maintenance, you can recover.
Keep in mind that maintenance can be disruptive to repository users. It is a good idea to communicate repository maintenance with your team or repository followers. Make sure everyone has checked in their code and have agreed to cease development during maintenance.
Understanding file removal from Git history
Recall that cloning a repository clones the entire history — including every version of every source code file. If a user commits a huge file, such as a JAR, every clone thereafter includes this file. Even if a user ends up removing the file from the project with a subsequent commit, the file still exists in the repository history. To remove this file from your repository you must:
- remove the file from your project's current file-tree
- remove the file from repository history - rewriting Git history, deleting the file from all commits containing it
- remove all reflog history that refers to the old commit history
- repack the repository, garbage-collecting the now-unused data using git gc
Git 'gc' (garbage collection) will remove all data from the repository that is not actually used, or in some way referenced, by any of your branches or tags. In order for that to be useful, we need to rewrite all Git repository history that contained the unwanted file, so that it no longer references it - git gc will then be able to discard the now-unused data.
Rewriting repository history is a tricky business, because every commit depends on it's parents, so any small change will change the commit id of every subsequent commit. There are two automated tools for doing this:
- the BFG Repo Cleaner - fast, simple, easy to use. Require Java 6 or above.
- git filter-branch - powerful, tricky to configure, slow on big repositories. Part of the core Git suite.
Remember, after you rewrite the history, whether you use the BFG or filter-branch, you will need to remove
reflog entries that point to old history, and finally run the garbage collector to purge the old data.
Using the BFG to rewrite history
The BFG is specifically designed for removing unwanted data like big files or passwords from Git repos, so it has a simple flag that will remove any large historical (not-in-your-current-commit) files: '--strip-blobs-bigger-than'
$ java -jar bfg.jar --strip-blobs-bigger-than 100M
Any files over 100MB in size (that aren't in your latest commit - because your latest content is protected by the BFG) will be removed from your Git repository's history. If you'd like to specify files by name, you can do that too:
$ java -jar bfg.jar --delete-files *.mp4
Alternatively, using git filter-branch to rewrite history
filter-branch command rewrites a Git repo's revision history, just like the BFG, but the process is slower and more manual. If you don't know where the big file is, your first step will be to find it:
Manually reviewing large files in your repository
Antony Stubbs has written a BASH script that does this very well. The script examines the contents of your packfile and lists out the large files. Before you begin removing files, do the following to obtain and install this script:
- Download the script to your local system.
- Put it in a well known location accessible to your Git repository.
Make the script an executable:
$ chmod 777 git_find_big.sh
- Clone the repository to your local system.
- Change directory to your repository root.
Run the Git garbage collector manually.
git gc --auto
Find out the size of the .git folder.
$ du -hs .git/objects 45M .git/objects
Note this size down for later reference.
List the big files in your repo by running the
$ git_find_big.sh All sizes are in kB's. The pack column is the size of the object, compressed, inside the pack file. size pack SHA location 592 580 e3117f48bc305dd1f5ae0df3419a0ce2d9617336 media/img/emojis.jar 550 169 b594a7f59ba7ba9daebb20447a87ea4357874f43 media/js/aui/aui-dependencies.jar 518 514 22f7f9a84905aaec019dae9ea1279a9450277130 media/images/screenshots/issue-tracker-wiki.jar 337 92 1fd8ac97c9fecf74ba6246eacef8288e89b4bff5 media/js/lib/bundle.js 240 239 e0c26d9959bd583e5ef32b6206fc8abe5fea8624 media/img/featuretour/heroshot.png
The big files are all JAR files. The pack size column is the most relevant. The
aui-dependencies.jarcompacts to 169KB but the
emojis.jarcompacts only to 580. The
emojis.jaris a candidate for removal.
You can pass this command a filter for rewriting the Git index. For example, a filter can remove a file from every indexed commit. The syntax for this is the following:
git filter-branch --index-filter 'git rm --cached --ignore-unmatch pathname' commitHASH
--index-filter option modifies a repo's staging (or index). The
--cached option removes a file from the index not the disk. This is faster as you don't have to checkout each revision before running the filter. The -
-ignore-unmatch option in
git rm prevents the command from failing if the pathname it is trying to remove isn't there. By specifying a commit HASH, you remove the
pathname from every commit starting with the HASH on up. To remove from the start, leave this off or you can specify HEAD.
If all your large files are in different branches, you'll need to delete each file by name. If all the files are within a single branch, you can delete the branch itself.
Option 1: Delete files by name
Use the following procedure to remove large files:
Run the following command to remove the first large file you identified:
git filter-branch --index-filter 'git rm --cached --ignore-unmatch filename' HEAD
- Repeat Step 1 for each remaining large file.
Update the references in your repository.
filter-branchcreates backups of your original refs namespaced under
refs/original/. Once you're confident that you deleted the correct files, you can run the following command to delete the backed up refs, allowing the large objects to be garbage collected:
$ git for-each-ref --format="%(refname)" refs/original/ | xargs -n 1 git update-ref -d
Option 2: Delete just the branch
If all your large files are on a single branch, you can just delete the branch. Deleting the branch automatically removes all the references.
Delete the branch.
$ git branch -D PROJ567bugfix
Prune all of the reflog references from the branch on back.
$ git reflog expire --expire=now PROJ567bugfix
Garbage collecting dead data
Prune all of the reflog references from now on back (unless you're explicitly only operating on one branch).
$ git reflog expire --expire=now --all
Repack the repository by running the garbage collector and pruning old objects.
$ git gc --prune=now
Push all your changes back to the Bitbucket repository.
$ git push --all --force
Make sure all your tags are current too:
$ git push --tags --force