backup.py 文件源码

def client_cleanup(self):
        self.job.update_state(BackupJob.State.client_done, BackupJob.State.client_cleanup)

        del self.job.client_manifest_data
        del self.job.client_manifest_id_str
        try:
            del self.job.client_key_data
            del self.job.client_key_type
        except AttributeError:
            pass

        transaction.get().note('Deleted client keys of job %s' % self.job.id)
        transaction.commit()
        # TODO delete checkpoints

        # TODO do we actually want this? if we leave the cache, the next job has a good chance of rsyncing just a delta
        # TODO perhaps a per-client setting, to limit space usage on the client with multiple repositories.