I have suse leap 15 with root placed in btrfs partition, after I was installed webcamoid packet, my system is damaged (perhaps it related with my special nvidia driver). I try to delete some last snapshot and rollback to previous state of my system, but result is unsuccessful (because snapper has lost correct root point it was lost a functionality), and now my system not loading.
Now after booting my system halted only in grub2 menu. Suse rescue disk gui also wrote ‘no linux system’.
But if I look to mounted my damaged btrfs partition, I see in /@/boot folder a number of files like - vmlinuz-xxxxxx, vmlinux-xxxxxx, initdr-xxxxxx, system.map-xxxxxx, system-conf-xxxxxx, symtypes-xxxxxx, symvers-xxxxxx, sysctl.conf-xxxxxx, config-xxxxxx and other. Therefore my first questions - what command I can use in suse rescue console to load my system from this existing image?
And second question - I have full DD-copy of my harddisk in ftp-server, what command I can use in suse rescue console to restore my harddisk from ftp-server?
This is different copy of my system, first may be damaged after intallation webcamoid, however in ftp-server I have not so actual, but fully correct DD-copy of my system.
Thanks, Viacheslav
I don’t know if this helps. I have avoided “btrfs”, and so I have little experience with this.
Mount the root partition at “/mnt”. (Or mount it else and adapt the other commands to fit).
Then:
mount --bind /dev /mnt/dev
mount --bind /proc /mnt/proc
mount --bind /sys /mnt/sys
You still need to mount the “btrfs” subvolumes.
The easiest way to do this seems to be:
chroot /mnt
mount -a
exit
The “exit” there is just from the chroot shell. You can skip that and continue in the chroot shell if you want.
The “mount -a” in the chroot shell should have mounted what else is needed, including “btrfs” subvolumes, EFI partition (if used), etc.
I have no experience repairing “btrfs” systems beyond that. But at this stage you should have everything mounted. I also don’t know much about the specific problems that you are having. But if that mount attempt for the subvolumes works, then you can probably make more progress than in your first attempts.
thank you for your help, I mount btrfs partitution many times successfully with subvolume or without, but my problem is further and deeper - how I can reboot from image what been placed in this subvolume or how I can fully replace this btfrs subvolume from ftp.
You seem to have a way to make backups of your system which you describe a bit vague like a “dd of the system” (probably a dd of the root file sytem to a file?) and that you put on another sytem using ftp.
I assume that that when you designed that way of backup, you also designed the way to restore. because those two always have to be done together.
From your vague description, I would say, ftp the file from your backup system back to your rescue system on a place somewhere there (of course not within the root file sytem of the broken system, but either on a free place somewhere on the disk of the broken system, or on an USB device) and dd back from there.
But as that is the very obvious thing to do, I guess there is more behind your question then I read from it.
BTW, I do not use btrfs, but your last line in the above post is not something you can do. You can not restore a sub-volume on it’s own. A sub-volume is only a usable thing together with the whole btrfs file system. You have to restore the complete contents of the file sytem (partition in your case).
Please take a look at the following SDB article: <https://en.opensuse.org/SDB:BTRFS>.
There was a presentation at the 2017 openSUSE Conference on the topic of “Btrfs and it’s maintenance and repair” but, it seems that access to the 2017 Conference Schedule is currently blocked …
If all that fails to resolve the issue, you’ll have to reinstall your system’s Boot, UEFI and System partitions – there’s no need to reinitialise the User (XFS) partition.