Lack of locking mechanism allow concurrent backups
Using duplicity handler and remote storage (cloud based for instance), when backing up large sets in particular for initial or periodic full backups, it takes a long time to perform a backup run, sometimes longer than 24 hours.
The subsequent backup execution will try to execute a cleanup, which will mess with the running backup.
This is no good.
A running backup should prevent subsequent ones to start, with some locking mechanism.
This has been exhibited in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=511300
(from redmine: created on 2010-12-24, closed on 2010-12-26)