Hi,
This problem has already destroyed my system three times*. When I was upgrading or installing a package, during pre-installation hooks, timeshift-autosnap deleted the snapshot the system had booted into, crashing it (It doesn’t happen all the time, but at seemingly random times):
I’ve now disabled autosnap to prevent it from breaking my system again. Is there anything I can do so that autosnap doesn’t delete the snapshot the system is booted to?
Tanks a lot
simdll
*with a live EndeavorOS I managed to modify the boot script from grub and boot the system and then restore grub.
When you boot off a snapshot, you should never install or remove packages. That will likely put you in a very bad place where you are permanently booting into your snapshot.
When you a boot a snapshot, the only thing you should do is restore the snapshot, not use the system.
This is a correct configuration, or I need to fix it? I have set autosnap to left 3 snapshot, delete the oldest and update grub.
Thank you for the support
simdll
The problem isn’t your settings. You are just doing something you shouldn’t be doing. When you boot off a read/write snapshot, you must not install/remove/update packages.
I must set boot options directly from /@/boot? When I boot in to a snapshot I selected the last one and I try to upgrade the system and system crashed. This is the problem, right? Setting the boot from the normal path I should back in a normal situation?
Thank you
Yes. This is the issue. This scenario isn’t intended to work. Booting a snapshot is so you can have a working session to restore a snapshot.
If the boot is currently wrong, one way to fix it is to go into /boot/grub/grub.cfg and manually edit the kernel options to the correct subvolume. Although if the system crashed, it is perhaps not an issue. You would need to check.
