Given that the rebase is a major upgrade, there is always a good chance that the upgrade will break your system. If your system is critical, then the very best way to do an upgrade it to test it first on a clone of your system, not your active system.
Given how dirt cheap small portable ssds and thumbnail drives are, and assuming you were wise and have your home folder on a separate partition, your system partition should be less that 64gigs in size.
This simple step that takes 15 minutes, may save you hours of heartache.
I do request that the devs and moderators of Neon and this forum include instructions about how to properly backup and restore your system with a clone tool such as clonezilla.
Timeshift and other “system restore” tools may not be able to recover from a major upgrade because too many things have been replaced that may not be covered by their images.
Those of you who use Samsung drives should perhaps use Samsung’s own clone tool.
With the last rebase, I tested on a clone for a couple of weeks before committing to the actual upgrade, which went perfectly as I had discovered and fixed all the things that were wrong with my active base system BEFORE jumping into the deep end.
The harsh reality of those who are having problems is their systems are usually not actually properly setup in the first place. Many people do non-standard things, force setups the wrong way, are using long depreciated methods, or do major customizations, then come and complain about updates breaking things.
As I stated in the Kubuntu facebook group I admin there is an old coder saying: “garbage in, garbage out”.
Yeah. I said the same thing here and here. But do they listen? Nope.
I even suggested using the GUI version of Clonezilla, Rescuezilla. Then I restore the image to a virtual machine to test how the upgrade goes. If there is anything I need to fix as far as dependencies, I don’t upgrade. This has never failed me.
And yes. I also put the user profile on another partition when it comes to any computer I have. I don’t keep anything on the OS partition. It’s beneficial that way also should I ever want to install a different KDE distribution. And I am thinking of it with what Canonical has been doing to Ubuntu.
Rather than Clonezilla I tend to use BTRFS snapshots (or Timeshoft rsync snapshots on non-BTRFS systems).
Do a snapshot, do the upgrade, roll back the snapshot if stuff breaks or delete it if you don’t have problems for a week.
That way you need less space and both snapshotting and rolling back are really simple and quick with Timeshift.
I agree 100% with you. And I use exactly your method.
I always do a ghost copy of my systems (Linux & Windows) before doing any “major” upgrades. I even had a discussion with someone here on this forum arguing that doing a HDD or system partition copy (Ghost or Clonezilla) is not the best way to do and it’s better to use Timeshift. But the 2 methods have their purpose and clearly Timeshift doesn’t help here. I’m laughing now. For my needs I always do a “Ghost” copy and it saved me lot of hours of annoyances.
Indeed. Timeshift has pulled my behind out of the fire more than once. Unfortunately with this level of upgrade, Timeshift tends to be limited and I have had it fail. There is still too much risk, which is why I do clones. That way there is no risk.