Hi, we have an Amanda 3.3.1 server on Debian Wheezy that is in production since a couple of months now.
Today I was quite surprised to see the amdump email complaining for every entry in the disklist file for one of the three backup sets we have, filled with error messages like this:

FAILED [can't dump in non degraded mode]

It looks like nobody else had a similar problem until now. However, after some research, I tried to dump again and this time amdump used next virtual tape and everything was fine. So it looks like that for some reason, one of our tape directories has been broken (tape number 23, to be more specific). How is that possible, being it a just a directory on the server disks?
Moreover, should I just replace that virtual tape, removing the directory and running amlabel -f on it?

Thanks in advance for your hints,

-- Davide