Glitches on new install and Backup/Restore

I just reinstalled my system with 4.1.2 and restored qubes with the backup. I’ve noticed a couple of weird glitches - I wonder if any of you have noticed something similar?

  1. Restore Backup seemed to work but failed to finish at the end. After leaving it for a couple of hours, I hit cancel and got an error message, but everything seemed ok and there was no temporary restore file in /etc/var (or /var/etc, or whatever it uses).

  2. One App-vm was restored in name only - none of my Firefox bookmarks, history or extensions have been copied across. Its like a blank disp-vm (There could be more of these, i haven’t checked yet).

  3. My e-mail vm has restored, but the timestamp in Thunderbird is now 14 hours behind.

(Also having problems getting redshift to read its config file, but I think that’s a different issue.)

Just wondering if these sorts of glitches are common? Particularly unhappy with an error-prone backup/restore, I have to say.

1 Like

/var/tmp, but in any case cancelling a restore is almost guaranteed to result in missing data. It can take many hours to complete depending on the backup size, and there’s an unintuitive aspect in case of very sparse VM volumes: E.g. restoring a VM with only 5 GB actually used but “Private storage max size” set to 500 GB in Qube Settings will take much longer than restoring one with 5 GB used that had the max set to 20 GB.

Try running the qvm-backup-restore CLI for just that one VM, maybe there’s some interesting diagnostic message.

Did you shut down the e-mail VM before backing it up?

Thanks for this reply.

I don’t see documentation for the commandline, so I have to use the GUI.

I deleted the offending qube and restored it from backup again, letting it run for a long time until its “Finished successfully” message. I clicked ok.

Strangely, the Backup Restore dialog window stayed open, behind everything else. It showed the following:

-> Restoring comp-issues...
Extracting data: 12.2 GiB to restore
-> Done.
-> Please install updates for all the restored templates.

The progress bar was full/complete. The Finish button was grayed out, only the Cancel button was available. I cancelled, but got a warning from dom0, something like the window might be busy and is not responding. Do you want to terminate the application? Yes / No. I terminated it.

The qube seems to be working properly now.

I tried the same approach with the email qube (delete then restore again). Firstly, it hasn’t done anything for the timestamp, still around 14-15h out of wack.

Secondly, the dialog box stayed open again in the same way. This time, I could not close it with cancel - it just seemed frozen/non-responsive. Multiple clicks on the top right X generated the window might be busy... dialog. Again, I chose to close it.

BTW, the email qubes was indeed closed when I took a backup of it.

There’s some available with qvm-backup-restore --help or man qvm-backup-restore, although it’s not a lot.

To restore two qubes named foo and bar from a backup file /mnt/removable/qubes-backup-2023-04-05T123456 in sys-usb, the command would be:

qvm-backup-restore -d sys-usb /mnt/removable/qubes-backup-2023-04-05T123456 foo bar

That really does sound like a bug worth reporting in the issue tracker. Maybe someone can help diagnose it there:

When you wrote this I assumed you meant that 14 hours worth of previously downloaded mail were missing. Now I’m not sure if that’s right? (If you just meant that the system time is wrong, there were some recent threads in the forum about fixing time synchronization.)

6 posts were split to a new topic: Thunderbird timezone and date formats