I'm in the midst of upgrading an old iMac Mid 2011. By now I upgraded to High Sierra. I bought an SSD disk and cloned it with an USB drive. I could not use Mac's clone function: there was a read error after which all stopped. I used ultracopier to do the Clone process. After about 15 hours, all was copied with 1 IO read error.
After some investigating I found out that a Windows 10 guest's *.pvm file was the problem. Which is weird, because the guest boots just fine on the old HDD. Disk Utilities' EHBO did not found any problems on the old HDD either (but SMART logs confirm the disk desperately needs replacement).
I then went to remove old Snapshots. I could delete a few recent ones, but Snapshots made with Parallels 9 (by me now upgraded to Parallels v15) cannot be deleted with the user interface.
After some Googling, I kinda figured out how the virtual machines work with Mac OS's parallels. SO I moved (copy file, delete file) 2 old Snapshots out of the Windows10.pvm (show archive contents) - Windows10-0.hdd - Windows10-0.hdd.0{STRING}.hds. Sure enough, a read error occured when I tried to copy the oldest .hds file out of the .pvm. Still, I continued and removed the 2 files, including the corrupt one. I tried to start the Windows10 guest-file from the old HDD. It wouldn't boot. So I made a copy of DiskDescriptor.xml (in Windows10-0.hdd) and did an edit, removing the blocks that referenced the 2 .hds files. Parallels still wouldn't boot (The hard disk image file used by the hard disk 6 (0) is corrupt and its data cannot be recovered).
So now I'm stuck - what to do?? I'll be moving the 2 .hds files back into the Windows10-0.pvm and restore the original DiskDescriptor.xml and hope at least the old HDD guest will boot again.
While I await confirmation of that: HELP - WHAT SHOULD I DO??
After some investigating I found out that a Windows 10 guest's *.pvm file was the problem. Which is weird, because the guest boots just fine on the old HDD. Disk Utilities' EHBO did not found any problems on the old HDD either (but SMART logs confirm the disk desperately needs replacement).
I then went to remove old Snapshots. I could delete a few recent ones, but Snapshots made with Parallels 9 (by me now upgraded to Parallels v15) cannot be deleted with the user interface.
After some Googling, I kinda figured out how the virtual machines work with Mac OS's parallels. SO I moved (copy file, delete file) 2 old Snapshots out of the Windows10.pvm (show archive contents) - Windows10-0.hdd - Windows10-0.hdd.0{STRING}.hds. Sure enough, a read error occured when I tried to copy the oldest .hds file out of the .pvm. Still, I continued and removed the 2 files, including the corrupt one. I tried to start the Windows10 guest-file from the old HDD. It wouldn't boot. So I made a copy of DiskDescriptor.xml (in Windows10-0.hdd) and did an edit, removing the blocks that referenced the 2 .hds files. Parallels still wouldn't boot (The hard disk image file used by the hard disk 6 (0) is corrupt and its data cannot be recovered).
So now I'm stuck - what to do?? I'll be moving the 2 .hds files back into the Windows10-0.pvm and restore the original DiskDescriptor.xml and hope at least the old HDD guest will boot again.
While I await confirmation of that: HELP - WHAT SHOULD I DO??