Hi all,
we're facing little annoyance with our tape backup. Our tape library (Quantum SuperLoader3) apparently has no option to run through all available LTO-9 tapes and calibrate/initialize them. What we can do is load them manually using the library's management tool one by one to have them calibrated, but since the tape changing takes some minutes and the calibration takes about 1h, this is annoying and time consuming.
It would be cool if PBS could just wait for such a calibration to finish when it loads a tape to barcode label it, because that is something that can work as a batch. I've seen that there was apparently a patch integrated some time ago to address this: https://lists.proxmox.com/pipermail/pbs-devel/2025-April/013121.html
But either I'm doing something wrong, or it doesn't work with our library. Here's what happens when we have some unlabeled and uncalibrated tapes in the library and then start the "Barcode label" operation:
It can be seen that the load is started and times out after about 12 minutes. And since the drive is still calibrating, all subsequent commands fail. I'll see if it recovers at least when the drive becomes idle; sometimes this state even requires a reset of the tape library to restore communication with PBS.
We had this issue with PBS 3.4 and I've updated it to 4.1 this week, but this apparently didn't change. Is there a way to simply expand the tape loading timeouts to e.g. 2h, so that PBS has patience to wait until the drive finishes?
Thanks!
we're facing little annoyance with our tape backup. Our tape library (Quantum SuperLoader3) apparently has no option to run through all available LTO-9 tapes and calibrate/initialize them. What we can do is load them manually using the library's management tool one by one to have them calibrated, but since the tape changing takes some minutes and the calibration takes about 1h, this is annoying and time consuming.
It would be cool if PBS could just wait for such a calibration to finish when it loads a tape to barcode label it, because that is something that can work as a batch. I've seen that there was apparently a patch integrated some time ago to address this: https://lists.proxmox.com/pipermail/pbs-devel/2025-April/013121.html
But either I'm doing something wrong, or it doesn't work with our library. Here's what happens when we have some unlabeled and uncalibrated tapes in the library and then start the "Barcode label" operation:
Code:
2026-03-06T15:50:40+01:00: checking/loading media 'WA3542L9'
2026-03-06T16:02:46+01:00: unable to load media 'WA3542L9' - load drive failed - Aborted Command, Additional sense: Logical unit communication time-out
2026-03-06T16:02:46+01:00: checking/loading media 'WA3543L9'
2026-03-06T16:08:11+01:00: scsi command failed: transport error
2026-03-06T16:08:11+01:00: unable to load media 'WA3543L9' - error reading element status: read element address assignment failed: scsi command failed: transport error
2026-03-06T16:08:11+01:00: checking/loading media 'WA3544L9'
2026-03-06T16:08:14+01:00: Unit Attention, Additional sense: Power on, reset, or bus device reset occurred
2026-03-06T16:13:39+01:00: unable to load media 'WA3544L9' - error reading element status: read element status (B8h) failed: scsi command failed: transport error
2026-03-06T16:13:39+01:00: media 'WA3546L9' already inventoried (already labeled)
2026-03-06T16:13:39+01:00: checking/loading media 'WA3547L9'
2026-03-06T16:13:42+01:00: Unit Attention, Additional sense: Power on, reset, or bus device reset occurred
2026-03-06T16:13:43+01:00: Unit Attention, Additional sense: Not ready to ready change, medium may have changed
2026-03-06T16:19:07+01:00: unable to load media 'WA3547L9' - error reading element status: read element status (B8h) failed: scsi command failed: transport error
2026-03-06T16:19:07+01:00: checking/loading media 'WA3548L9'
2026-03-06T16:19:09+01:00: Unit Attention, Additional sense: Power on, reset, or bus device reset occurred
2026-03-06T16:19:10+01:00: Unit Attention, Additional sense: Not ready to ready change, medium may have changed
2026-03-06T16:24:34+01:00: unable to load media 'WA3548L9' - error reading element status: read element status (B8h) failed: scsi command failed: transport error
2026-03-06T16:24:34+01:00: checking/loading media 'WA3549L9'
2026-03-06T16:24:37+01:00: Unit Attention, Additional sense: Power on, reset, or bus device reset occurred
2026-03-06T16:24:38+01:00: Unit Attention, Additional sense: Not ready to ready change, medium may have changed
2026-03-06T16:30:02+01:00: unable to load media 'WA3549L9' - error reading element status: read element status (B8h) failed: scsi command failed: transport error
2026-03-06T16:30:02+01:00: TASK WARNINGS: 14It can be seen that the load is started and times out after about 12 minutes. And since the drive is still calibrating, all subsequent commands fail. I'll see if it recovers at least when the drive becomes idle; sometimes this state even requires a reset of the tape library to restore communication with PBS.
We had this issue with PBS 3.4 and I've updated it to 4.1 this week, but this apparently didn't change. Is there a way to simply expand the tape loading timeouts to e.g. 2h, so that PBS has patience to wait until the drive finishes?
Thanks!