Tapes are no longer written full after host change

[mkretzer]

Hello,

ticket 05859792.

Until 3 weeks ago our LTO8 tape drives was attached to our old Windows Server Core system. Now we attached the drives to our W2022 Veeam server.

Since then all drives no longer write the tape until full, some are written full, but some are only written 1 TB, 3 TB and so on. There are some strange messages coming from the tape according to support (Tape prepare error : Tape error: '6' (The handle is invalid.) , Tape error: '1' (Incorrect function.)).
Upgrading drivers did not help (which was our biggest hope).

Does anyone know what that can be? It is very unlikely that all 4 drives now suddenly have a fault.

Markus

[Mildur]

Hi Markus

I also don't expect all 4 drives to be faulted just because you moved them to a new host. That would be really bad luck.
Have you tried a diagnostic tool of your tape vendor to check if the tape still reports the entire size?

I suggest continuing working with our support team to figure this out. Maybe a case with your tape vendor would also be helpful. Let's see what our support engineer recommends next after analyzing the new log package.

Best,
Fabian

[mkretzer]

Hello,

i'll start our core system again and reattach the drives. Then support will compare logs.

Markus

[mkretzer]

An update: i have been working on this with Veeam & Quantum. Test results:

- The issue still does not happen on the old server. All tapes are always written 100 % full.
- I did copy a tape that has only been written up to 800 GB and then did a full write test with IBM tape tools attached to the "problematic" server to the original tape - there was no issue and the tape could be written 100 % full.
- While this is going on i am constantly do verifications on all the tapes from the "problematic" and "good" server - all the tapes were written 100 % good

It really seems Veeam has a problem writing these tapes on the new server!