The backup_client role implements backups using duply/duplicty. Each backup involves not only the data itself, but also a lot of metadata that accompany such backups. Over time these metadata files tend to occupy quite a bit of space on both the backup server and the backup client without providing any value (since the will start to fall outside of the designated backup retention policies).

It would be useful to automatically get rid of any excess fails accumulated during backup, thus reducing disk usage, and making the system run more smoothly in general.

The following should be done:

• Update the backup_client role.
  • Add running purge --force at the end of backup process. This should purge the metadata files which are past the retention policy preserve times.
  • Run the pre/bkp/post/purge (with --force) stages in sequence in order to ensure that each step is executed successfully.
• Update role reference documentation.
• Update usage instructions.
• Update release notes.