The issue happened on a fresh Arch Linux installation but the solution here can be used on any systemd-based distributive.
So, during restarts system hangs for 1.5 minutes with a message like:
A stop job is running for Session c1 of user setevoy (1min 30s)
To check which process(es) exactly caused this you can use journalctl with the -p5 (–priority) option:
[simterm]
$ journalctl -p5
[/simterm]
Then scroll down till the end and find messages containing “Killing”:
Feb 25 20:04:14 setevoy-arch-pc systemd[1]: session-1.scope: Stopping timed out. Killing.
Feb 25 20:04:14 setevoy-arch-pc systemd[1]: session-1.scope: Killing process 20420 (sudo) with signal SIGKILL.
Feb 25 20:04:14 setevoy-arch-pc systemd[1]: session-1.scope: Killing process 20429 (openvpn) with signal SIGKILL.
Feb 25 20:04:14 setevoy-arch-pc systemd[1]: session-1.scope: Failed with result ‘timeout’.
Next just need to find out why exactly those processes won’t stop in a normal way.
Another way is just to decrease timeouts to force stop such a process.
To do so – edit the /etc/systemd/system.conf file and set DefaultTimeoutStartSec and DefaultTimeoutStopSec options which are has 90 seconds by default:
... DefaultTimeoutStartSec=30s DefaultTimeoutStopSec=10s ...
Reload systemd‘s configs and on a next reboot they will be killed after 10 seconds timeout on shutdown:
[simterm]
$ sudo systemctl daemon-reload
[/simterm]
Done.
