Strange Tape Logic

[IanBolton]

I have been using VBR for a few years and never noticed this...

I am restoring a few VMs from tape to repo as validation. The selected restore point is an incremental one, so the VBKs and VIBs are on different tapes. But I believe all of the VIBs are on the same tape and the VBKs are on the same tape.

So when I monitor the restore session, I see the full tape and the daily tape switching after each vib/vbk per VM. This seems mad - why not restore all of the files required from one tape, then from another, etc. Or utilise a second drive and restore from both at once? The way its doing it seems to add time for multiple loads/unloads and wear on the robotics/drives!

Code: Select all

Message
Loading tape 003010L8 from Slot 5 to Drive 1 (Server: CONA...)
Tape drive Tape2 is ready
Restoring file Alpha-vm-358980D2026-03-25T223420...
Unloading tape 003010L8 from Drive 1 (Server: CONA...)
Loading tape 003420L8 from Slot 15 to Drive 1 (Server: CONA...)
Tape drive Tape2 is ready
Unloading tape 003420L8 from Drive 1 (Server: CONA...)
Loading tape 003010L8 from Slot 5 to Drive 1 (Server: CONA...)
Tape drive Tape2 is ready
Unloading tape 003010L8 from Drive 1 (Server: CONA...)
Loading tape 003420L8 from Slot 15 to Drive 1 (Server: CONA...)
Tape drive Tape2 is ready

[david.domask]

Hi IanBolton,

Quick question, how many drives are available for the media pool?

Similarly, are you doing a restore to Repository or Restore to Infrastructure?

The workflow is a bit different depending on the restore method, so please share a few more details, then will be able to offer better guidance, thanks!

[IanBolton]

there are three in the Library... Not sure on the pool - if I look at the pool it's assigned to the Library and sees three drives..
The restore is to repo (I will then selectively IR them to prove the backups good)

Its not a huge issue and so probably not worthy of bothering support I was just curious.

Ian