Tumbleweed System Broken

English Install/Boot/Login
1

My Suse Tumbleweed is broken.
I had not run it in a while and when I did, it wanted to do updates, over 3 GB of downloads, so I let it. But it got stuck on 74% and after I waited for 4 hours, it was still stuck. Then I saw it was frozen. No response to anything. Would not shut down or respond to any command, nor control-alt-del or a 1 second push of the power button. I had to shut off the power. I booted from a sysrec usb stick and ran e2fsck. The output is included below. Now when I try to boot Tumbleweed it does not work. I get the login screen, select my user and enter the password, and then the system is frozen again. Not as bad as the first time: I can get it to shut down with a 1 second push of the power button. I can also boot the tumbleweed system in recovery mode, in command line as root - but I don’t know what to do there. Any suggestions on how I might recover the system?

I have tried e2fsck with -vpf and vpfcc - the output looks like this:

e2fsck -vpf -C 0 /dev/sda2
Inode 655934 extent tree (at level 1) could be shorter. IGNORED.
Inode 8267686 extent tree (at level 1) could be shorter. IGNORED.
1104731 inodes used (6.74%, out of 16384000)
1861 non-contiguous files (0.2%)
1366 non-contiguous directories (0.1%)
# of inodes with ind/dind/tind blocks: 0/0/0
Extent depth histogram: 818757/404
16912584 blocks used (25.81%, out of 65536000)
0 bad blocks
3 large files
713085 regular files
81548 directories
0 character device files
0 block device files
4 fifos
145902 links
310024 symbolic links (285498 fast symbolic links)
60 sockets

 1250623 files
2

I updated a Tumbleweed system yesterday, that had not been used for 6 months.

At several points during the update, it reported an error. It gave me a choice of retry,ignore,abort. I went with ignore. This seemed to happen about 20 times during the update.

After the update, I rebooted. And I then ran “zypper dup” again. This time it update 20 packages – presumably the ones that had failed on the first run. And all seems okay.

1 Like
3

What I usually do is run zypper dup -d until all packages download (I’ve just gotten in the habit of doing a pre-download before running zypper dup to apply the updates.

That way if there are issues with the download, I can do a zypper ref && zypper dup -d in case I’ve caught something mid-sync on the mirrors.

1 Like
4

You can get into failsafe mode. You can also update from the console in text mode with "sudo zypper dup "

5

Are you able to boot from a read-only snapshot?

6

Apparently zypper will not work in the rescue mode command line. It produces more than a screen of error messages. I think that it is not finding any repositories or may not be on-line, but I can’t handle all that technical stuff. Android mentioned booting from a snapshot: how would one do that? Apparently I have a snapshot directory with what appears to be a copy of the system. But I don’t know if it is possible to boot from a directory.

What would happen if I just mounted the drive and copied the actual files
to overwrite / ?

7

How to get internet connection in command line:

8

Don’t copy files to recover.

Recovery using snapshots is done from the grub menu - System recovery and snapshot management with Snapper | Reference | openSUSE Leap 15.6 is written for Leap, but is the same for Tumbleweed.

9

I tried
zypper dup -y
and it worked.
The system is running again. Thanks.
I did not know you could hit F10 to boot the command line.
They might put that on the screen.

10

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.