etckeeper is a collection of tools to let
/etc be stored in a git,
mercurial, bazaar or darcs repository. This lets you use git to review or
revert changes that were made to
/etc. Or even push the repository
elsewhere for backups or cherry-picking configuration changes.
It hooks into package managers like apt to automatically commit changes
made to /etc during package upgrades. It tracks file metadata that git does
not normally support, but that is important for /etc, such as the
It's quite modular and configurable, while also being simple to use if you understand the basics of working with version control.
First, a big warning: By checking /etc into version control, you are creating a copy of files like /etc/shadow that must remain secret. Anytime you have a copy of a secret file, it becomes more likely that the file contents won't remain secret. etckeeper is careful about file permissions, and will make sure that repositories it sets up don't allow anyone but root to read their contents. However, you also must take care when cloning or copying these repositories, not to allow anyone else to see the data.
Since git mushes all the files into packs under the .git directory, the whole .git directory content needs to be kept secret. (Ditto for mercurial and .hg as well as bazaar and .bzr)
Also, since version control systems don't keep track of the mode of files like the shadow file, it will check out world readable, before etckeeper fixes the permissions. The tutorial has some examples of safe ways to avoid these problems when cloning an /etc repository.
Also note that
etckeeper init runs code stored in the repository.
So don't use it on repositories from untrusted sources.
What etckeeper does
etckeeper has special support to handle changes to /etc caused by
installing and upgrading packages. Before apt installs packages,
etckeeper pre-install will check that /etc contains no uncommitted changes.
After apt installs packages,
etckeeper post-install will add any new
interesting files to the repository, and commit the changes.
You can also run
etckeeper commit by hand to commit changes.
There is also a cron job, that will use etckeeper to automatically commit any changes to /etc each day.
Version Control Systems are designed as a way to manage source code, not as a way to manage arbitrary directories like /etc. This means there are a few limitations that etckeeper has to work around. These include file metadata storage, empty directories, and special files.
Most VCS, including git, mercurial and bazaar have only limited tracking of
file metadata, being able to track the executable bit, but not other
permissions or owner info. (darcs doesn't even track executable bits.) So
file metadata is stored separately. Among other chores,
sets up a
pre-commit hook that stores metadata about file owners and
permissions into a
/etc/.etckeeper file. This metadata is stored in
version control along with everything else, and can be applied if the repo
should need to be checked back out.
git and mercurial cannot track empty directories, but they can be
significant sometimes in /etc. So the
pre-commit hook also stores
information that can be used to recreate the empty directories in the
Most VCS don't support several special files that you probably won't have
in /etc, such as unix sockets, named pipes, hardlinked files (but symlinks
are fine), and device files. The
pre-commit hook will warn if your /etc
contains such special files.
Darcs doesn't support symlinks, so they are also stored in
A quick walkthrough of using etckeeper.
Note that the default VCS is git, and this tutorial assumes you're using it. Using other VCSes should be broadly similar.
First, get etckeeper installed. Something like:
apt-get install etckeeper
etckeeper init command initialises an /etc/.git/ repository.
If you installed etckeeper from a package, this was probably automatically
performed during the package installation. If not, your first step is to
run it by hand:
etckeeper init command is careful to never overwrite existing files
or directories in /etc. It will create a
.gitignore if one doesn't
already exist (or update content inside a "managed by etckeeper" comment
block), sets up pre-commit hooks if they don't already exist, and so on. It
does not commit any files, but does
git add all interesting files for
an initial commit later.
Now you might want to run
git status to check that it includes all
the right files, and none of the wrong files. And you can edit the
.gitignore and so forth. Once you're ready, it's time to commit:
cd /etc git status git commit -m "initial checkin" git gc # pack git repo to save a lot of space
After this first commit, you can use regular git commands to handle further changes:
passwd someuser git status git commit -a -m "changed a password"
Rinse, lather, repeat. You might find that some files are changed by daemons and shouldn't be tracked by git. These can be removed from git:
git rm --cached printcap # modified by CUPS echo printcap >> .gitignore git commit -a -m "don't track printcap"
etckeeper hooks into apt (and similar systems) so changes to interesting
files in /etc caused by installing or upgrading packages will automatically
be committed. Here "interesting" means files that are not ignored by
You can use any git commands you like, but do keep in mind that, if you check out a different branch or an old version, git is operating directly on your system's /etc. If you do decide to check out a branch or tag, make sure you run "etckeeper init" again, to get any metadata changes:
git checkout april_first_joke_etc etckeeper init
Often it's better to clone /etc to elsewhere and do potentially dangerous
stuff in a staging directory. You can clone the repository using git clone,
but be careful that the directory it's cloned into starts out mode 700, to
prevent anyone else from seeing files like
fixes their permissions:
mkdir /my/workdir cd /my/workdir chmod 700 . git clone /etc cd etc etckeeper init -d . chmod 755 ..
Another common reason to clone the repository is to make a backup to a
server. When using
git push to create a new remote clone, make sure the
new remote clone is mode 700! (And, obviously, only push over a secure
transport like ssh, and only to a server you trust.)
ssh server 'mkdir /etc-clone; cd /etc-clone; chmod 700 .; git init --bare' git remote add backup ssh://server/etc-clone git push backup --all
If you have several machines using etckeeper, you can start with a etckeeper repository on one machine, then add another machine's etckeeper repository as a git remote. Then you can diff against it, examine its history, merge with it, and so on. It would probably not, however, be wise to "git checkout" the other machine's branch! (And if you do, make sure to run "etckeeper init" to update file permissions.)
root@darkstar:/etc>git remote add dodo ssh://dodo/etc root@darkstar:/etc>git fetch dodo root@darkstar:/etc>git diff dodo/master group |head diff --git a/group b/group index 0242b84..b5e4384 100644 --- a/group +++ b/group @@ -5,21 +5,21 @@ sys:x:3: adm:x:4:joey tty:x:5: disk:x:6: -lp:x:7:cupsys +lp:x:7:
Incidentially, this also means I have a backup of dodo's /etc on darkstar. So if darkstar is compromised, that data could be used to attack dodo too. On the other hand, if dodo's disk dies, I can restore it from this handy backup.
Of course, it's also possible to pull changes from a server onto client machines, to deploy changes to /etc. Once /etc is under version control, the sky's the limit..
The main configuration file is
etckeeper runs the executable files in
ignores the same ones that run-parts(1) would ignore.) You can modify these
files, or add your own custom files. Each individual file is short, simple,
and does only one action.
For example, here's how to configure it to run
git gc after each apt run,
which will save a lot of disk space:
cd /etc/etckeeper/post-install.d (echo '#!/bin/sh' ; echo 'exec git gc') > 99git-gc chmod +x 99git-gc git add . git commit -m "run git gc after each apt run"
Here's how to disable the automatic commits after each apt run, while still letting it git add new files:
etckeeper will notice if it's being run by way of sudo, and makes a commit
with the author set to the user who sudoed to root. This is useful when
a system has multiple admins; as long as they use sudo while doing their
administration, and run
etckeeper commit to commit their changes,
git blame can show who was responsible for each change.
By default, etckeeper uses git. This choice has been carefully made; git is the VCS best supported by etckeeper and the VCS users are most likely to know.
[ It's possible that your distribution has chosen to modify etckeeper so its default VCS is not git -- if they have please complain to them, as they're making things unnecessarily difficult for you, and causing unnecessary divergence of etckeeper installations. You should only be using etckeeper with a VCS other than git if you're in love with the other VCS. ]
If you would like to use some other VCS, and
etckeeper init has already
been run to set up a git repository, you have a decision to make: Is the
history recorded in that repository something you need to preserve, or can
you afford to just blow it away and check the current /etc into the new
In the latter case, you just need to follow three steps:
etckeeper uninit # deletes /etc/.git! vim /etc/etckeeper/etckeeper.conf etckeeper init
In the former case, you will need to convert the git repository to the
other VCS using whatever tools are available to do that. Then you can
etckeeper uninit, move files your new VCS will use into place,
etckeeper.conf to change the VCS setting, and finally
etckeeper init. This procedure is clearly only for the brave.
Two blog posts provided inspiration for techniques used by etckeeper:
isisetup had some of the same aims as etckeeper, however, unlike it, etckeeper does not aim to be a git porcelain with its own set of commands for manipulating the /etc repository. Instead, etckeeper provides a simple setup procedure and hooks for setting up an /etc repository, and then gets out of your way; you manage the repository using regular VCS commands.
etckeeper is licensed under version 2 or greater of the GNU GPL.
Joey Hess email@example.com