When the server returns a temporary failure message, which is to be expected some times when restoring data to a loaded cluster, the cbrestore tool prints out a generic error message, like:
Error on key 'session:t43o4t8aa116etmh9r89ras9v2': Memcached error #134: Temporary failure
It then sleeps for a second, and then retries the operation. That part is fine. This bug is just to print something nicer for the end user so it is clear that the item is being retried and not lost, and that everything is functioning as designed.
For example, it could write something like:
2012-06-12 10:55:33 Temporary failure when writing key 'session:t43o4t8aa116etmh9r89ras9v2'. Server is loaded. Retrying in 1 second....
2012-06-12 10:58:33 Saw 15 Temporary failures in the past 3 minutes