ss operates using a user config file (.ss_config) stored in a directory, from which the ss command can be run. In technical terms ss only sources the ".ss_config" file in the directory from which ss is invoked. This config file can be created using "ss createconfig". One can then use "ss initial" to create an initial full backup. This backup is stored on the destination in a directory called "0". After the "ss initial" command is run once, one can repeatedly call "ss incremental" (using cron for example) to create incremental snapshots (ss uses rsync's --link-dest option to create hard links). This creates directory's on the destination starting from the initial full backup directory "0", so the first incremental snapshot directory will be called "1", the next one "2" etc. Use "ss list" to retrieve a listing of all the snapshots made on the destination, including dates and transfer sizes that are sed'd from rsync by this script.
ss stores the snapshot date and other info in a file called "snapshot.info" in the snapshot directory. "ss list" uses this information to create a listing.
ss stores specific (non-user config) in a file called ".ss_vars", in the same directory where ".ss_config" is stored. In this version only one variable is used to keep track of the snapshot directory numbers.
I used rsync's --link-dest option to facilitate this requirement by creating hard links. Using hard links, all the incremental backups are transparent as seen from the user, and minimal space is used. All previous snapshots are passed to rsync's --link-dest option to create hardlinks whenever possible.
Not being able to (easily) retrieve backups in case of system failure defeats the purpose of creating backups in the first place. rsync's --link-dest option takes care of this already, to create transparent snapshots. This includes the possibility to just delete snapshots on the destination, hard links make this operation transparent.
Copy the ss bash script somewhere in your $PATH, or add it's location to your (user) $PATH. Make it executable using chmod u+x ss or chmod ugo+x ss and run one of the following command line options explained below.
ss without any options displays all the available command line options.
ss creates a default config file named .ss_config in the current directory. Adjust as needed (self explainatory).
ss produces a list of all the snapshots made in a file list.ss in the current directory.
Create the first initial backup. This option also creates a file .ss_vars used for housekeeping.
Create snapshot. If the backup process is interrupted and ss incremental is run again, the previous interrupted backup on the destination will be overwritten.
- GNU sed