Arvin Schnell, 10 May 2017
So far the user had to ensure that snapshots created by rollbacks get deleted to avoid filling up the storage. This process has now been automated. During a rollback snapper sets the cleanup algorithm to “number” for the snapshot corresponding to the previous default subvolume and for the backup snapshot of the previous default subvolume.
Here is a simple example. We start with a newly installed system. Snapshot #1 is the system and btrfs default subvolume. Snapshot #2 is a snapshot created by YaST right after the installation:
# snapper --iso list
Type | # | Pre # | Date | User | Cleanup | Description | Userdata
-------+---+-------+---------------------+------+---------+-----------------------+--------------
single | 0 | | | root | | current |
single | 1 | | 2017-05-09 10:05:13 | root | | first root filesystem |
single | 2 | | 2017-05-09 10:10:42 | root | number | after installation | important=yes
# btrfs subvolume get-default /
ID 259 gen 244 top level 258 path @/.snapshots/1/snapshot
Now we do a rollback:
# snapper rollback
Creating read-only snapshot of default subvolume. (Snapshot 3.)
Creating read-write snapshot of current subvolume. (Snapshot 4.)
Setting default subvolume to snapshot 4.
After the rollback the previous system (snapshot #1) and the backup snapshot of the previous system (snapshot #3) have “number” set as cleanup algorithm:
# snapper --iso list
Type | # | Pre # | Date | User | Cleanup | Description | Userdata
-------+---+-------+---------------------+------+---------+-----------------------+--------------
single | 0 | | | root | | current |
single | 1 | | 2017-05-09 10:05:13 | root | number | first root filesystem |
single | 2 | | 2017-05-09 10:10:42 | root | number | after installation | important=yes
single | 3 | | 2017-05-09 12:17:30 | root | number | rollback backup of #1 | important=yes
single | 4 | | 2017-05-09 12:17:30 | root | | |
As you can see the snapshots #1 and #3 have the cleanup algorithm set thus the cleanup algorithm will automatically remove them as soon as the cleanup limits are reached. Snapshot #3 also has the important flag set so it should be keep longer in case the user wants to copy data from it. Snapshot #4 is the new default btrfs subvolume and thus the system after a reboot.
This feature is available in snapper since version 0.5.0 and only for btrfs.