Manages tarsnap backups
A wrapper around tarsnap which does two things:
$ apt-get install python-setuptools $ easy_install tarsnapper
A configuration file looks like this:
# Global values, valid for all jobs unless overridden: # A job's delta controls when old backups are expired # (see "How expiring backups works" below) deltas: 1d 7d 30d # You can avoid repetition by giving deltas names delta-names: super-important: 1h 1d 30d 90d 360d # A job's target sets the name of the created archive target: /localmachine/$name-$date # You can also include jobs from separate files include-jobs: /usr/local/etc/tarsnapper/*.yml jobs: # define a job called images (names must be unique) images: source: /var/lib/mysql exclude: /var/lib/mysql/temp exec_before: service mysql stop exec_after: service mysql start # Aliases can be used when renaming a job to match old archives. alias: img some-other-job: sources: - /var/dir/1 - /etc/google excludes: - /etc/google/cache target: /custom-target-$date.zip deltas: 1h 6h 1d 7d 24d 180d
For the images job, the global target will be used, with the name placeholder replaced by the backup job name, in this case images.
You can then ask tarsnapper to create new backups for each job:
$ tarsnapper -c myconfigfile make
The name of the archive will be the target option, with the $date placeholder replaced by the current timestamp, using either the dateformat option, or %Y%m%d-%H%M%S.
Or to expire those archives no longer needed, as per the chosen deltas:
$ tarsnapper -c myconfigfile expire
If you need to pass arguments through to tarsnap, you can do this as well:
$ tarsnapper -o configfile tarsnap.conf -o v -c tarsnapper.conf make
This will use tarsnap.conf as the tarsnap configuration file, tarsnapper.conf as the tarsnapper configuration file, and will also put tarsnap into verbose mode via the -v flag.
Using the include-jobs option, you could insert 1 or more jobs in (for example) /usr/local/etc/tarsnapper/extra-backup-jobs.yml:
# Included jobs act just like jobs in the main config file, so for # example the default target is active and named deltas are # available, and job names must still be globally unique. yet-another-job: source: /var/dir/2 deltas: 1h 1d 30d an-important-job: source: /var/something-important delta: super-important
include-jobs uses Python’s globbing to find job files and hence is subject to the limitations thereof.
If you want to create the backups yourself, and are only interested in the expiration functionality, you can do just that:
$ tarsnapper --target "foobar-\$date" --deltas 1d 7d 30d - expire
The --target argument selects which set of backups to apply the expire operation to. All archives that match this expression are considered to be part of the same backup set that you want to operate on.
tarsnapper will then look at the date of each archive (this is why you need the $date placeholder) and determine those which are not needed to accommodate the given given delta range. It will parse the date using the python-dateutil library, which supports a vast array of different formats, though some restrictions apply: If you are using yyyy-dd-mm, it cannot generally differentiate that from yyyy-mm-dd.
You can specify a custom dateformat using the --dateformat option, which should be a format string as expected by the Python strptime function (e.g. %Y%m%d-%H%M%S). Usually, a custom format is not necessary.
Note the single “-” that needs to be given between the --deltas argument and the command.
The expire command supports a --dry-run argument that will allow you to see what would be deleted:
$ tarsnapper --target "foobar-\$date" --deltas 1d 7d 30d - expire --dry-run
The design goals for this were as follows:
The generations are defined by a list of deltas. 60 means a minute, 12h is half a day, 7d is a week. The number of backups in each generation is implied by it’s and the parent generation’s delta.
For example, given the deltas 1h 1d 7d, the first generation will consist of 24 backups each one hour older than the previous (or the closest approximation possible given the available backups), the second generation of 7 backups each one day older than the previous, and backups older than 7 days will be discarded for good.
The most recent backup is always kept.
As an example, here is a list of backups from a Desktop computer that has often been running non-stop for days, but also has on occasion been turned off for weeks at a time, using the deltas 1d 7d 30d 360d 18000d:
dropbox-20140424-054252 dropbox-20140423-054120 dropbox-20140422-053921 dropbox-20140421-053920 dropbox-20140420-054246 dropbox-20140419-054007 dropbox-20140418-060211 dropbox-20140226-065032 dropbox-20140214-063824 dropbox-20140115-072109 dropbox-20131216-100926 dropbox-20131115-211256 dropbox-20131012-054438 dropbox-20130912-054731 dropbox-20130813-090621 dropbox-20130713-160422 dropbox-20130610-054348 dropbox-20130511-055537 dropbox-20130312-064042 dropbox-20120325-054505 dropbox-20110331-121745