document how to restore files as root (hopefully)

    $ ssh -tt bacula-director-01.torproject.org bconsole

    *restore

... and follow instructions. Reminder: by default, backups are

restored on the originating server. `llist jobid=N` and `messages` to

follow progress.

... and follow instructions. 

Reminder: by default, backups are restored to the originating server,

as the `bacula` user, in `/var/tmp/bacula-restores`. Make sure that

directory has enough disk space, and if you need to restore specific

ownership settings, you want to run `bacula-fd` as root, see the

[restoring files as root](#restoring-files-as-root) section.

Use the `llist jobid=N` and `messages` commands to follow progress.

The `bconsole` program has a pretty good interactive restore mode

which you can just call with `restore`. It needs to know which "jobs"

See the [upstream manual](https://www.bacula.org/9.4.x-manuals/en/main/Restore_Command.html) more information about the [restore

command](https://www.bacula.org/9.4.x-manuals/en/main/Restore_Command.html).

### Restoring files as root

Note that the above procedure restores files as the `bacula` user,

which means it cannot create files owned by another user. For basic

use cases that doesn't matter: files can be changed owner by hand

after restore.

But if you're restoring a full system or more complex configuration

than just a single service, that will likely not work because you'd

need to manually change files individually, and that doesn't scale.

So the solution is to run `bacula-fd` as root, temporarily. Follow

this procedure:

 1. Disable Puppet

        systemctl stop puppet-run.timer

 2. Remove the configuration file that tells `bacula-fd` to run as the

    `bacula` user:

        rm /etc/systemd/system/bacula-fd.service.d/bacula-fd-groups.conf

 3. Reload the daemon:

        systemctl daemon-reload

        systemctl restart bacula-fd

 4. On the director, run the backup (follow the above procedure):

        bconsole

        [...]

 5. Files in `/var/tmp/bacula-restores` (or wherever you put them)

    should now have the right owner

 6. Reset Puppet:

        pat

The last step will restart the systemd timer, reset the configuration

file, and restart all daemons (in theory).

TODO: Note that this procedure was never tested, but this, in theory,

works. Please confirm and update this section after success or failure.

## Restore a host that has been offline for a long time

If a host has been offline for a long time its storage configuration might have been expired by puppet. You will notice that when you are trying to restore some file to a different host you will get the following error after having selected the files: