Hello, I remove xterm, it’s dependency icewm and can’t load now even in console mode (ctrl+alt+f1). I loaded from read-only snapshot, enter the console, add both packets via yast but it didn’t helps. So I try to upgrade my leap via usb image. I found that my Kde and plasma disappears from packages menu (was unchecked) so I’m added it. But it didn’t helps too. So, I tried to recover by snapper’s snapshots in yast (I use Btrfs) and recover one that was before my xterm uninstall actions. Nothing helps…
You probably removed all of X, or almost all of X. I’m not sure what command you used to remove “xterm”, but it probably gave a long list of other software that would also be removed.
I’m honestly not sure how to recover from this. You can try booting to single user mode. I think that’s a ‘1’ at the end of the grub boot line.
If that works, then you should be able to run yast at the terminal, to reinstall a whole bunch of stuff.
Failing that, try a reinstall. Perhaps tell the installer to upgrade your existing system, as the easiest way to keep as much as possible of what you have.
You can try a repair using a DVD Install disk.
Then, follow that up with a “zypper up” to bring your system back up to date.
TSU
Don’t have an optical drive more than 5 years. My biggest USB Flash drive is 4 GB and I use network openSUSE installer. I goes into recovery from it but don’t know what to do with it. It loads console with recovery root profiler but there are no external commands in bash like yast, zipper, sudo etc.
Can I recover with BTRFS snapshots or rollback last yast actions?
I remove it via Yast GUI. So, the list was not too long, about 5-8 packages I think.
In the GRUB load options I have only:
- Load OpenSuse
- load specific kernel
- load read-only snapshot
need to check another options and chance to edit the loader
For 2) although the screen likely chopps off the description at the right margin,
Each kernel selection should be a pair…
The first time the kernel is listed would boot normally with that kernel.
The second time that kernel is listed would boot into recovery mode, which <might> work if your system is not heavily damaged, you would have networking plus basic drivers and functions. Anything deemed “extra” would likely be disabled.
If you can at least boot to a console in recovery mode, you could try a “distribution upgrade”
sudo zypper dup
If you want to see the entire GRUB entries, I wrote scripts to fix, but this is something you might do in the future after you fix your system.
I haven’t updated them for LEAP (I probably should), but the 13.2 package can be used…
You just need to open the scripts in a text editor, follow the instructions in the comments and modify as needed… Specifically not pointing the background image to the included 13.2 image.
https://en.opensuse.org/User:Tsu2/12.3/Modified_GRUB_Menu
TSU
If the “zypper dup” try doesn’t work or is not possible,
And you aren’t willing to get a USB device large enough to hold the DVD image to do a repair,
Then IMO
You should consider a re-install using your USB net installer,
The critical steps you would need to make to preserve your /home partition
- Identify your /home partition by its partition id.
- Know where you keep your personal data and files. Most will be in your /home, but if there are any in your root partition, you need to make copies elsewhere.
- Being very careful not to make a mistake, remove the swap and root partitions. Making a mistake here is irrecoverable. For this reason, you may want to mount your disk on a LiveCD or other running OS and copy your /home partition to other media.
- Run your openSUSE install. It should find your unpartitioned free space and offer to install a default layout. DECLINE THE OFFER. Edit the proposed layout, pointing the proposed /home partion to your still existing /home partition. Only then accept the changes. and proceed with the re-installation.
- When re-installation is complete, test everything to make sure they’re working.
- Then update your machine
zypper update
7 Install apps and add additional repos as needed.
TSU
As nrickert already wrote, you likely removed more than that.
Of course you cannot fix things when booting from a read-only snapshot. It’s read-only after all.
But you can switch your system back to that read-only snapshot after booting it, by running:
sudo snapper rollback
See also https://doc.opensuse.org/documentation/leap/reference/html/book.opensuse.reference/cha.snapper.html#sec.snapper.snapshot-boot
Btw, IIRC, grub2 should have told you that when you selected the snapshot… 
I recovered via USB installer GUI. Just add some packets to X. Need to remove extras later.
Also I don’t understand - is snapshot recover is effective? In yast it showed me 3 dirs with changes: /home, /etc and sometimes extra one. But it didn’t recover my packets (or filesystem clusters or what it should recover via LVM). Need to read more about it.
Hmm, I watch my snapshots in KDE GUI now and see that they contains many folders and changes and should be very effective.
My xterm packet deletes something important in this dependencies. Should check the choosen one:
xdm-1.1.11-13.1.x86_64 xinit-1.3.4-2.4.x86_64
icewm-1.3.12-3.2.x86_64
xdg-menu-0.2-217.2.noarch
xorg-x11-essentials-7.6_1-17.2.noarch
xdm-xsession-1.1.11-13.1.x86_64
fvwm2-2.6.6-2.4.x86_64
icewm-default-1.3.12-3.2.x86_64
icewm-lang-1.3.12-3.2.noarch
xorg-x11-7.6_1-17.2.noarch
xosview-1.16-6.4.x86_64
xorg is the hart of the GUI system you drove a stack through it. :’(
So, icewm can be leleted safely? 
No
Removing any desktop is not safe since it may remove stuff needed for other desktops.
Just don’t run it. Besides Icewm is a nice emergency desktop that generally works if your main desktop gets broken for some reason
ok, thanks all. Next subj - switch to TW or wait LE 42.3
Actually it should be safe to uninstall icewm, at least if you don’t want to use VNC (it will remove xorg-x11-Xvnc too which requires icewm).
But yes, it is a nice “emergency” desktop, or if you want to save as much RAM as possible.
And it’s quite small, so I don’t see why you would want to remove it.
Of course it is effective.
The “snapper rollback” should revert your system to exactly the same state as when the snapshot is created.
Some directories are excluded from snapshots by default though, for good reason (e.g. to avoid data loss when doing a full roll back).
E.g. /home/, /tmp/, /var/lib/mariadb/
But it didn’t recover my packets (or filesystem clusters or what it should recover via LVM). Need to read more about it.
No, it only saves/restores files.
But packages consist of files, and the RPM database (that contains the information what packages are installed) consists of files too.
If you only revert single files, the RPM database is not changed though.