1

Running xubuntu 18.04 LTS the boot animation does not work anylonger (instead I see a black screen) and it takes very long to boot. I changed /etc/default/grub the line #GRUB_CMDLINE_LINUX_DEFAULT="quiet splash" to #GRUB_CMDLINE_LINUX_DEFAULT="" and after a sudo update-grub I do see boot messages on the screen. At a certain point presumably X server or lightdm is starting and the screen blacks out as before. After 2-3 minutes the login screen appears normally.

I don't know what caused this problem, bug I think some update or configuration change by me, but I don't remember anything relevant. I've tried the properitary Nvidia driver and the default one. I use two monitors.

/var/log/lightdm/lightdm.log shows a huge gab of >2min between two lines:

...
[+0.94s] DEBUG: Session pid=1104: Authentication complete with return value 0: Success
[+0.94s] DEBUG: Seat seat0: Session authenticated, running command
[+0.94s] DEBUG: Launching process 1107: xubuntu-numlockx
[+1.03s] DEBUG: Process 1107 exited with return value 0
[+1.03s] DEBUG: Seat seat0: Exit status of xubuntu-numlockx: 0
[+1.03s] DEBUG: Session pid=1104: Running command /usr/lib/lightdm/lightdm-greeter-session /usr/sbin/lightdm-gtk-greeter
[+1.03s] DEBUG: Creating shared data directory /var/lib/lightdm-data/lightdm
[+1.03s] DEBUG: Session pid=1104: Logging to /var/log/lightdm/seat0-greeter.log
[+130.54s] DEBUG: Activating VT 7
[+130.54s] DEBUG: Activating login1 session c1
[+130.54s] DEBUG: Seat seat0 changes active session to c1
[+130.54s] DEBUG: Session c1 is already active
[+130.81s] DEBUG: Greeter connected version=1.26.0 api=1 resettable=false
[+131.18s] DEBUG: Greeter start authentication for peter
[+131.18s] DEBUG: Session pid=1585: Started with service 'lightdm', use
...

I've added allow-debugging = true to /etc/lightdm/lightdm-gtk-greeter.conf which adds some debug messages to /var/log/lightdm/seat0-greeter.log but AFAICS the messages start after the mysterious two minute sleep.

I couldn't figure out what happens between those two lines. Any idea how I could debug this further?

Edit: the problems disappeared, I don't know how.

  • An alternate approach is open terminal and use journalctl and page down through the boot log for exact program name that is taking too long. – WinEunuuchs2Unix Aug 12 '18 at 14:19
2

Your lightdm login greeter may be hanging due to not having enough entropy. See https://unix.stackexchange.com/questions/442698/when-i-log-in-it-hangs-until-crng-init-done.

In my case, this issue started appearing when upgrading to the latest kernel on Debian sid. I had to wait for the following kernel message in order to continue logging in.

crng init done
7 urandom warnings missed due to ratelimiting

Adding more entropy to the kernel by using haveged or rng-tools, fixes the login issue.

  • No idea about OP, but this DID fix it for me (on ArchLinux). – Darkhogg Jan 15 '19 at 21:18
  • There should be a systemd unit systemd-random-seed.service whose job it is to save 512 random bytes during shutdown in /var/lib/systemd/random-seed and to feed that back as startup entropy into the kernel random driver during the next boot, to prevent exhaustion of the randomness pool. If an exhausted entropy pool really is the problem, you should see during this hang (on a separate console) that /proc/sys/kernel/random/entropy_avail is close to zero (should normally be closer to ~4000). See man 4 random for details. – Markus Kuhn Aug 9 '19 at 16:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.