I do believe what happened to you has happened to me. I read that thread many times and I just couldn’t figure it out. I kept googling “Timeshift restore” and never got a clue on how to restore a snapshot correctly. At the moment, without Google and just going through the process of “restore”, I would swear I’m all done… but the truth is, I’m making things worse.
I took screenshots of the process because Timeshift only ever told me two things.
After opening up the Timeshift GUI, selecting a snapshot and hitting restore…
- Warning Data will be modified. Please save your work and close all applications. System will reboot after files are restored (but the system never reboots like the rsync version).
- After hitting next, the last prompt I see: Restore completed. Restored subvolumes will become active after system is restarted. You can continue working on the current system. After restart, the current system will be visible as a new snapshot…
Here is what I thought would/should happen.
Reboot and be done with it.
But, what actually happens is, silently my Grub has changed to boot me into one specific snapshot. On first boot, it does look like the restore worked. Without any more prompts, notices, suggestions otherwise, I think I’m all done but the truth is, it’s a sinister time bomb.
Why? Did you know booting into a snapshot you could get caught up in Timeshift cleaning that snapshot out from under you? Seriously. One moment, it’s all good and the next, your apps start to lock up, your wallpaper is gone, no mouse or keyboard input and nothing left to do… It’s as if it’s all being erased before your eyes. You hard boot the computer and realize that snapshot is now gone because you just can’t boot the system up…
This happened to me 3 or 4 times before I went into the Timeshift schedule and made every option 9999.
Although that doesn’t really fix anything, it stopped Timeshift from cleaning out the snapshot I might be currently working with.
The last update (a few days ago) completely broke my Pipewire sound server. Pipewire crashed almost all the time and caused the system to lag and choke. I only knew of one option to quickly roll back and hope for the best. So, I pulled up Timeshift and hoped for the best. I restored an update and it seemed to work but what I didn’t know was, I was about to get stuck in a loop of frustration and cursing the heavens.
After restore, I thought I was free to experiment but it just so happens, I can’t seem to restore from a restore that wasn’t committed yet? I would go through the process to restore but upon reboot, I’m right back in a broken system. Timeshift is saying “restore complete” but after a reboot, it’s just as broken as how I left it. It was so bad I spent several hours figuring out Snapper. In the process of figuring out Snapper I discovered how to truly restore with Timeshift.
But, does it work? Yeah. Does it work the way I imagine it? Yeah. When it works its great but you just need to finish it up. No one told me this. I never got a clue about it and I had to figure it out on my own from a competitive product. I’m so deep in with Snapper at this point I might as well go all the way. But, I love Timeshift and it was reliable with rsync. I recommended it to everyone and I believe in it.
So, I think, Timeshift on Manjaro with BTRFS must be broken. We’re not being told to “finish” anything. If you make a restore with Timeshift and follow my first post, you’re good. After that, Timeshift becomes reliable. But it just doesn’t work this way right out of the box. I’m thinking it must be a bug.
