[Info-vax] Problems restoring image-backup

[David Froble]

hb wrote:
> On 06/28/2016 06:25 PM, Michael Moroney wrote:
>> jouk <joukj at hrem.nano.tudelft.nl> writes:
> 
>>> $ mount/over=(id,lim) dka100
>>> %MOUNT-F-NOHOMEBLK, Files-11 home block not found on volume
>>> -MOUNT-I-VOLIDENT, label = '............', owner = '............',
>>> format = '...
>>> .........'
>> Does a BACKUP/IMAGE restore create GPT.SYS on the destination volume when 
>> done on an Itanium?  ($ HELP BACKUP /GPT suggests no)
>> An Alpha boot loader wouldn't know what to do with GPT.SYS.
>> Look to see if [000000]GPT.SYS exists on the supposedly bootable volume.
> 
> Usually there is no [000000]GPT.SYS on an Alpha system disk. So a
> BACKUP/IMAGE restore should not restore that file. Restore also
> shouldn't create one. As far as I understand this file is created at
> init time. Which leads to the question, which actual commands were used
> to restore the image backup? Also the command to create the backup may
> be of interest. It is in the backup save set (if a save set was used at
> all). A BACKUP/LIST will show it in the summary section (or the bcksum
> utility will print just that summary). It also contains the backup
> version as well as the VMS version.
> 
> If you want to see what is on the restored disk, which you can't mount
> as an ODS disk, you should mount it /FOREIGN and dump the first two
> blocks. There are strings which identify the I64 (system) disks and the
> EFI/GPT.
> 
> 

My understanding is in an /IMAGE restore that all volume info comes from the 
save set.  Perhaps this is not correct?  If not, then I'd consider it a bug.