ClearLinux Boot Logs - Login Issue

Q&A
1

I’m having problems booting/logging into ClearLinux so I’m trying to find the relevant logs to help solve the issue. I would appreciate some help finding with which log files to look at to help solve this, or if anyone has any ideas what the problem might be I would appreciate any suggestions.

Issue

When the login screen appears my user icon is not there, and when I try to login my credentials work but it just restarts the login screen and brings me back to the user selection screen to login again. If I enter invalid credentials it won’t let me login because the credentials are invalid and does not restart the login screen. I had issues in the past where customizations made to the ~/.profile ~/.bashrc and some bash aliases caused issues logging in, so I removed those files and tried to boot again but still could not login.

Background

I see in /var/log there are a few files, but I don’t see anything that’s been helpful. I have a second installation of ClearLinux on a separate drive that I am using (I keep it for troubleshooting issues like this) but I do need to get the other installation working as I have spent a lot of time setting it up and it has some projects that I have put a lot of time into (things that can’t easily be transferred, they have to be installed and configured).

I have recently removed an external hard drive bay because a drive died, so I can’t just plug it back in to see if that would make it boot. I’m thinking there could be a file the computer is trying to read on the external bay that can’t; I have read other people who have had issues logging in have had missing files that caused the issue.

Recent Changes:

I have docker on the ClearLinux installation with some compose projects that I’ve been working on lately. I also had recently setup NFS and tried setting up Samba (but that didn’t work, I’m not sure if that’s still trying to start and failing, but I can’t find where that would be).

I used find to make a log of files that were modified in the last 10 days, but it’s fairly large and I can’t see anything that stands out. If someone could recommend some log files to look at, or some files to check that may be causing the issue, I would be very much appreciate it. Most of the changed files seem to be docker or system updates (I can’t tell which version I’m using because I can’t login).

Samba

I’m not sure if this is relevant but the /var/log/samba/log.samba file has the following error:

exit_daemon: daemon failed to start: Samba detected misconfigured 'server role' and exited. Check logs for details, error code 22

Logs

Here are the /var/log files:

-rw-rw----  1 root          utmp              3072 Feb  7 19:48 btmp
-rw-r--r--  1 root          root               219 Jan 17 19:21 dnf.librepo.log
-rw-r--r--  1 root          root              1654 Jan 17 19:21 dnf.log
-rw-r--r--  1 root          root                54 Jan 17 19:21 dnf.rpm.log
drwxr-xr-x  2 elasticsearch elasticsearch     4096 Jan 16 21:06 elasticsearch
drwx--x--x  2 root          gdm               4096 Jan 16 11:47 gdm
drwxr-xr-x  2 root          root              4096 Jan 16 18:03 httpd
drwxr-sr-x+ 3 root          systemd-journal   4096 Jan 16 11:47 journal
-rw-rw-r--  1 root          utmp            292584 Feb  6 20:20 lastlog
drwx------  2 root          root              4096 Mar 22 14:25 letsencrypt
drwxr-x---  2 httpd         httpd             4096 Mar  9 14:33 nginx
drwxr-x---  2 httpd         httpd             4096 Jan 16 18:03 nginx-mainline
drwx------  2 root          root              4096 Jan 15 21:43 private
drwx------  3 root          root              4096 Mar 24 09:42 samba
-rw-rw-r--  1 root          utmp             90624 Apr  1 09:16 wtmp
-rw-r--r--  1 root          andrew           31356 Apr  1 09:16 Xorg.0.log
-rw-r--r--  1 root          andrew           31356 Apr  1 09:16 Xorg.0.log.old

There is nothing in the gdm folder or private folder.

More Information

I made a couple gists if anyone thinks these might be helpful:

Thanks in advance

2

Can you login through another tty? (e.g `CTRL + ALT + F4)

This sticks out from your Xorg log
[ 30.591] (EE) systemd-logind: failed to get session: PID 3623 does not belong to any known session

If you can, take a look at the status of the systemd-logind service before and after your login attempt:
systemctl status systemd-logind.service

Otherwise take a look at the journal with journalctl -D.

1 Like
3

If you don’t find any breadcrumbs, you can try a repair:
https://docs.01.org/clearlinux/latest/guides/maintenance/fix-broken-install.html

The swupd repair command in that document probably needs updating to include --statedir=/mnt/var/lib/swupd nowadays.

Keep in mind that will replace and remove files under /usr. (not /usr/local though)

1 Like
4

Thanks @puneetse, I very much appreciate your quick responses. I took some time to try your suggestions but I still can’t log in.

I have tried the ctrl+alt+f4 at the login screen, it prompts me to login (saying <computer name> login:) but when I enter my username it says permission denied, and that’s the only user account I can log into. I can’t run systemctl since I can’t login; I can get no further in the terminal than the login prompt.
It seems that everything is fine until I try to login into the main user account, I was wondering if there would be a way to add a new administrator user account, maybe that would enable me to log in? I am guessing that something within my home directory is causing the issue since the OS is fine up until that point, but I am not sure.

I also tried the swupd repair with the --statedir however nothing was fixed. I didn’t use the --picky option because I have some manually installed manually that I was worried might be effected.

I would be very thankful for any thoughts, suggestions, or advice.
Thanks again

5

From a live image you can mount the rootfs of the Clear Linux that won’t let you login and run useradd with the --root parameter: sudo useradd --root=/mnt -G wheel puneetse

Or manually edit the /etc/passwd and /etc/group files under the mounted rootfs.

But if systemd-logind is actually failing like the log suggested, it still might not let you login. In which case you can try appending systemd.unit=rescue.target to your kernel boot parameters (hold SPACE during boot up to interrupt the bootloader and push e to edit the cmdline).

1 Like
6

I added a user, I had to use chroot with the useradd, the useradd with --root did not seem to work. Once I added a user I couldn’t set a password with passwd, just said:

passwd: Permission denied
passwd: password unchanged

I can’t seem to get to the kernal boot parameters by holding space while booting. Any other ideas or suggestions?