You are not logged in.
I've had an issue for ages that I cannot reproduce deterministically.
Once in 10-15 shutdowns systemd hangs (by default for 90 secs) then finally outputs a message in the spirit of the following before shutting down:
... session-c1.scope: Stopping timed out. Killing.
I produced a log following these instructions: https://freedesktop.org/wiki/Software/s … /#index2h1
Here are some highlights:
[ 3202.887803] systemd[1]: Received SIGCHLD from PID 917 (awesome).
[ 3202.887843] systemd[1]: Child 917 (awesome) died (code=exited, status=1/FAILURE)
[ 3202.887908] systemd[1]: session-c1.scope: Child 917 belongs to session-c1.scope
[ 3203.807735] systemd[1]: Received SIGCHLD from PID 720 (lxdm-binary).
[ 3203.807785] systemd[1]: Child 720 (lxdm-binary) died (code=exited, status=0/SUCCESS)
[ 3203.807878] systemd[1]: lxdm.service: Child 720 belongs to lxdm.service
[ 3203.807908] systemd[1]: lxdm.service: Main process exited, code=exited, status=0/SUCCESS
[ 3203.808264] systemd[1]: lxdm.service: Changed stop-sigterm -> dead
[ 3203.808793] systemd[1]: lxdm.service: Job lxdm.service/stop finished, result=done
[ 3203.808817] systemd[1]: Stopped LXDE Display Manager.
[ 3203.811700] systemd[1]: lxdm.service: cgroup is empty
[ 3203.811922] systemd[1]: lxdm.service: Collecting.
[ 3203.812036] systemd[1]: plymouth-quit.service: Collecting.
[ 3203.812046] systemd[1]: getty@tty1.service: Collecting.
[ 3203.812053] systemd[1]: getty.target: Collecting.
[ 3203.812186] systemd[1]: plymouth-quit-wait.service: Collecting.
[ 3203.817257] systemd[1]: Got cgroup empty notification for: /system.slice/lxdm.service
[ 3212.551751] systemd[1]: session-c1.scope: Stopping timed out. Killing.
[ 3212.551902] systemd[1]: session-c1.scope: Killing process 1213 (emacs) with signal SIGKILL.
[ 3212.552001] systemd[1]: session-c1.scope changed stop-sigterm -> stop-sigkill
As the log shows, I use LXDM, Awesome and Emacs (which is forked as a daemon).
Any lead?
Offline
It is back!
The annoying 90sec delay irregulary on shutdown/reboot with message
'A stop job is running for session c1...' since ca. a week.
Everything updated today.
Kernel: 4.11.5-1-ARCH
DE: openbox / lxqt
I think it might have been something to do with lxqt-panel.desktop or menu-cached?
# journalctl -b -1 (partial)
.
.
.
login[464]: pam_unix(login:session): session closed for user XXX
login[464]: pam_systemd(login:session): Failed to release session: Connection reset by peer
systemd-logind[402]: Failed to abandon session scope, ignoring: Transport endpoint is not connected
systemd[1]: Stopped Getty on tty1.
systemd[1]: Removed slice system-getty.slice.
lxqt-session[1]: lxqt-session: "Module 'lxqt-panel.desktop' won't terminate ... killing."
lxqt-session[1]: QProcess: Destroyed while process ("lxqt-panel") is still running.
lxqt-session[1]: The X11 connection broke: I/O error (code 1)
systemd[1]: session-c1.scope: Stopping timed out. Killing.
systemd[1]: session-c1.scope: Killing process 800 (menu-cached) with signal SIGKILL.
systemd[1]: Stopped Session c1 of user XXX.
systemd[1]: session-c1.scope: Unit entered failed state.
systemd[1]: Removed slice User Slice of .
systemd[1]: Stopping Login Service...
.
.
.
The number behind 'session c#' changes sometimes.
Can anyone tell me how to get rid of this 'A stop job is running...' again?
Offline
Moving to Applications and DEs.
Edit: Merging with an existing thread.
Note: roxosulaliwe's thread has been merged with this thread.
Last edited by x33a (2017-06-17 16:22:51)
Offline
@roxosulaliwe: You can reduce the timeout in /etc/systemd/system.conf, variable DefaultTimeoutStopSec.
Offline
I am trying to connect the dots of people experiencing more or less the same issue.
See my bug report and comments of others:
lxqt-panel: reboot and shutdown takes a long time
If you feel you have the same issue, and you believe lxqt is the issue, this might be a place to start.
Offline
I am not using lxqt, not mentioning the panel. I do use LXDM however. I'll try without it from now on and report back in a week or so.
Offline
OK, that wasn't it. LXDM is not the culprit (or at least not the only one).
Offline