Hello,
I have an LTO6 drive which sometimes seems to have some troubles, occasionally I get scsi command failed errors:
But this is not the problem I want to adress here. In the next job I get:
(Seems I have a problem with the power supply)
Now the only solution i to continue with the backups I have found is:
a) Set the Tape to full and continue with a new one
b) Delete the whole tape an redo all backups of that tape
Questions:
a) Do I oversee something? Is there an other solution?
b) Would it not be better, if the system would be more resilient to such errors? So that it can just continue the backup or redo the last job.
I see two solution to achive this:
1. Delete the old EOD only after the new EOD is written
2. If no EOD is found, Find the last unfinished job an redo it (I have no knowledge of the data structure on the tape but I assume this should be possible)
Thank You
I have an LTO6 drive which sometimes seems to have some troubles, occasionally I get scsi command failed errors:
Code:
2025-11-17T14:39:20+01:00: could not set encryption mode on drive: read data encryption caps SPIN(20h[0010h]) failed - do_scsi_pt failed with err ENODEV: No such device, ignoring.
2025-11-17T14:39:20+01:00: queued notification (id=5b30e6e2-9cdb-49c9-826a-e5bb2c587a13)
2025-11-17T14:39:20+01:00: TASK ERROR: write chunk failed - write failed - scsi command failed: transport errorBut this is not the problem I want to adress here. In the next job I get:
Code:
2025-11-18T14:00:08+01:00: TapeAlertFlags: TapeAlertFlags(DRIVE_VOLTAGE)
2025-11-18T14:00:08+01:00: moving to end of media
2025-11-18T14:06:09+01:00: queued notification (id=7eccfa99-fc08-4454-a595-ddae10d72007)
2025-11-18T14:06:09+01:00: TASK ERROR: move to EOD failed - Medium Error, Additional sense: End-of-data not found(Seems I have a problem with the power supply)
Now the only solution i to continue with the backups I have found is:
a) Set the Tape to full and continue with a new one
b) Delete the whole tape an redo all backups of that tape
Questions:
a) Do I oversee something? Is there an other solution?
b) Would it not be better, if the system would be more resilient to such errors? So that it can just continue the backup or redo the last job.
I see two solution to achive this:
1. Delete the old EOD only after the new EOD is written
2. If no EOD is found, Find the last unfinished job an redo it (I have no knowledge of the data structure on the tape but I assume this should be possible)
Thank You