System takes 90 seconds to reboot

I have been having an issue where my desktop take 90 seconds to reboot for a few weeks now. Just yesterday I removed quiet from my grub.cfg and saw the errors found in the attached images.

After looking around the web I saw someone mention it might be an issue with SDDM but I am not certain and could not find how exactly they solved it. I assume I would just switch to GDM or whatever else there is. I am not certain though if this is the issue because I have also had issues with the stability of plasmashell on Wayland. These could be two separate problems but I am not aware of how I would even test that.

For those who are more knowledgeable than me, what does this look like to you? Do I need to attach more information?

If it’s waiting for a user session to end then you have something running in the background when you shut down - check your processes and services. It could be something that requires network access but the network went down before the process finished.


You can reduce this default stop job shutdown value by editing /etc/systemd/system.conf


90s is far too long.


This is a good idea, but also be aware that things like profile-sync-daemon and anything-sync-daemon might need some time to sync files to disk on exit, especially with slower disks.


Here’s a tool that might help you find what process is hanging it up.
I’ve never tried it though so don’t know how useful it actually is

Affects me to. Escalated to:

Quickly switching to another terminal with ctrl+alt+F# and back is an easy and fast workaround.

1 Like

I edited the .conf and lowered the time to 15s and it still defaults to 90s. I went back and the file is saved correctly.

For that, check this stack exchange post.
If you do a ls /etc/systemd/system/ you’ll see the node name you’ll want to edit is display-manager.service Don’t edit the file directly. Read the entire post.

It would be better though to try and find out what is preventing your display manager from exiting properly instead of having systemd timeout and send a kill signal. But, I can understand not wanting to wait 90 seconds while troubleshooting :grin:

Please provide the output of the journal. You can get the journal from the previous boot/shutdown cycle by saying:

journalctl -b -1 --no-pager

1 Like

Did you uncomment those lines in addition to changing the values?