Tape backup job failed on tape 7 now tape is retired

[mike979]

I had done a full backup.

Was in the middle of the incremental when I lost power. The status was then

Code: Select all

9:28:32 PM	Data error (cyclic redundancy check). Tape positioning error: Tape error: '23' (Data error (cyclic redundancy check).)  --tr:Error code: 0x00000017 --tr:TapeStream initialization FAILED --tr:Failed to create MtfStream for read. --tr:Tape writer failed to append session. --tr:Cannot start new tape backup session 'Incremental backup set 4/17/2024 9:05:35 PM' for tape 'Tape0'. --tr:Cannot start new tape backup session 'Incremental backup set 4/17/2024 9:05:35 PM' for tape 'Tape0'. --tr:Client failed to process 	
9:28:42 PM	0 folders and 0 files backed up 	
9:28:42 PM	Error: Data error (cyclic redundancy check). Tape positioning error: Tape error: '23' (Data error (cyclic redundancy check).)  --tr:Error code: 0x00000017 --tr:TapeStream initialization FAILED --tr:Failed to create MtfStream for read. --tr:Tape writer failed to append session. --tr:Cannot start new tape backup session 'Incremental backup set 4/17/2024 9:05:35 PM' for tape 'Tape0'. --tr:Cannot start new tape backup session 'Incremental backup set 4/17/2024 9:05:35 PM' for tape 'Tape0'. --tr:Client failed to  	
9:28:42 PM	Job finished with error at 4/17/2024 9:28:42 PM 

When I restarted veeam, the tape was marked as retired.

1) Can I unretire this tape?

2) If not, if I put in a new tape and start the backup again, will it be smart enough to start over where the data was on tape 6? In other words, 5TB had already been written to this bad tape. If I use a new tape will the data on the bad tape be backed up on the new tape?

[david.domask]

Hi Mike,

Have you opened a Support Case yet? If not please open a Support case and be sure to include logs from the affected job as per https://veeam.com/kb1832 (use the first radio option and select the affected job)

Veeam will only retire tapes in a few situations where the drive returns an alert; I don't remember the exact list of alerts that may trigger it, but the point is more that if the tape got retired, it's because an alert was sent indicating.

It may be possible to unretire the tape depending on the circumstances, but it's best to let Support check the logs and see why the tape got retired; if it was because of a tape fatal error reported by the drive, likely it won't be usable.

For the job itself, it depends again on the nature of why the tape was retired, and there are a few too many "what-ifs" to list so it's best to let Support confirm the reason first and if the data on the tape is considered corrupted or not.

[mike979]

Yeah, it was a brand new tape, just lost power in the middle of the job.

I just ran a catalog and it came back with:

Code: Select all

4/17/2024 10:00:04 PM          Performing inventory for tape in Drive 1 (Server: TAPESERVER, Library: HPE Ultrium 8-SCSI, Drive ID: Tape0)
4/17/2024 10:22:59 PM Warning    Failed to read tape MoviesTape7 catalog
4/17/2024 10:22:59 PM          Error: Data error (cyclic redundancy check).                      
                               Tape positioning error: Tape error: '23' (Data error (cyclic redundancy check).)                               
4/17/2024 10:22:59 PM Error    Completed with error at Wednesday, April 17, 2024 10:22:59 PM

So given that the tape seems to be bad now, can I just re-run the backup job and will it start from where it left off?

[david.domask]

Hey Mike,

Got it, I definitely would open a Support Case as advised and let Support take a look; as noted, the call to retire is made by the drive alerts sent, so something likely isn't great here.

I can't tell precisely what the job will do as it's not clear what the full error was here; my guess is that the Tape Job considered that tape invalid and does not consider the restore points on the previous tape but I'm just guessing. You can check the tape properties by right-clicking on the tape in the UI and see if the restore points show there, and also under Backups > Tape on the Home tab of VBR.

Open a Support Case and share the case number here, and let's see what the Support team tells.

[Vodochnik]

Hi Mike,

to get the tape unretired in veeam you have to get it unretired in the library. If you look into the web interface of the library you will probably see your tape as retired too.
I managed to "unretire" a tape in an IBM library by removing a tape, let the library rescan, then reinserting a tape. Not just eject-reinsert a magazine, but eject, insert without "bad" tape, let library rescan, then eject a magazine and insert it with "bad" tape. Mailslot works too.

Veeam reads "bad" status from library.