I’ve just checked regarding symlinks pointing to tmpfs. That was indeed the cause. I was backing up /etc and there are symlinks there for /run and /dev. I’ve changed the backup include to be more specific, to /etc/cron.d/ and that has solved the issue for me.
Of course I’m still getting the ‘client calculated hash’ issue. There are no errors regarding snapshots?
The log today, shows:
Client calculated hash of “/backup/urbackup/puzzle/160926-0904/www/.datto_a6397fb09ab7afc8109af40ee4c432ff7f487e450b640935” differs from server calculated hash. This may be caused by a bug or by random bit flips on the client or server hard disk. Failing backup. (Hash: tree-sparse, client hash: bO8lWFUvGmy3ODP2WC4EkYVjmUig04r7mTvTv4glfwjHpNVlAPFiA5nqieK3JgM0pK5TLX5GYVeB33scX9CQyw==, server hash: tEGgfk51wQ4NvKLJhKzljp810l6zRV4vlIKLv95uJKrcDlvG//4UkO1JRybKqCgAH/a7zhBuynowy8h7cwOLWg==)
Fatal error during backup. Backup not completed
FATAL: Backup failed because of disk problems (see previous messages)
Backup failed
It looks to me like the snapshot was working properly this time? I could see the datto_snap_cow0 process was busy throughout the backup.
Do you have any suggestions for how to get the backup running in the meantime? Should I try removing and re-adding the client?