Recovering from a Bad Update¶
AXEM-SX uses Btrfs with snapper configured at install time.
Every zypper transaction creates a pre and post snapshot
automatically. That means: almost no system-level update is
truly fatal. You can roll back.
Read time: ~5 minutes.
This page is the calm reassurance article. Read it before you need it.
If the system still boots¶
Open a terminal:
```bash
List recent snapshots¶
sudo snapper list
Roll back to the last known-good snapshot (replace N with its number)¶
sudo snapper rollback N
Reboot¶
sudo reboot ```
After reboot, the system is exactly as it was at snapshot N. The broken transaction is undone.
If the system won't boot¶
At the GRUB menu, choose "Start bootloader from a read-only snapshot".
GRUB lists every available snapshot. Pick one from before the bad
update — typically the most recent one with a date earlier than your
last zypper call.
The system boots into that snapshot. Once you're in:
bash
sudo snapper rollback
sudo reboot
That promotes the snapshot you just booted to be the new running system.
Work in progress
A more complete version of this page will cover: per-/home
snapshots, manual snapshot creation before risky changes, GRUB
snapshot pruning, and the difference between snapper rollback
and zypper rollback.
What snapshots cover, and what they don't¶
| Covered | Not covered |
|---|---|
/ (the whole root filesystem) |
/home (by default — separate config) |
| Installed packages | External drives |
System configuration in /etc |
Browser profiles synced to a cloud |
| Kernel and bootloader changes | Files outside Btrfs subvolumes |
For documents and personal data, you still need a real backup. Snapper is for the system, not your work.
See also¶
- Snapshots & Recovery — index of the snapshots section.
- The Zypper Universe — why
zypperand snapper are designed to work together. - Your First Hour — read this before your first big update.
- openSUSE: Snapper — the upstream portal for snapper.
- snapper(8) manual page — the full command reference.