[Bug 1011053] Rebooting takes a long time because dmeventd refuses to exit (LVM thin volume)
http://bugzilla.suse.com/show_bug.cgi?id=1011053 http://bugzilla.suse.com/show_bug.cgi?id=1011053#c5 --- Comment #5 from Thiago Macieira <thiago@kde.org> ---
Does the problem happen every time when you shutdown or reboot? Or only occurred when the system is happening to run "dmeventd" deamon?
The two are the same thing because dmeventd is always running. My system is configured to snapshot /home every hour, so unless I start the system, do something and shut down before it runs, it'll happen.(In reply to zhen ren from comment #4)
What we need to known is what the lvm is doing there? We can imagine that snapper (hourly according to you) tell lvm to take snapshot, after that, the cleanup algorithm will also let lvm to cleanup the old snapshot. I don't know how long each of the operations lvm will take, which decides how busy lvm is, and the time window the shutdown meets a working lvm. It is reasonable if shutdown happens during lvm is busy. Anyway, I will look into more when I get time;-)
As far as I can tell: * hourly, snapper wakes up and tells lvm to take a snapshot * lvm starts dmeventd because of that * dmeventd NEVER exits * since dmeventd never exits, the D-Bus service fails to shut down in systemd, requiring a timeout before SIGKILL -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply@novell.com