ACPI (shutdown from web) not working only on one VM

Beyoogurt

New Member
Feb 21, 2015
3
0
1
Any idea, why ACPI shutdown from web not functioning on KVM VM?
It's working on every VM except one, which was migrated here by rsync (cloned from Digital Ocean).
All of them are Ubuntu 14.04.2 TLS


  • Package acpid already installed, tried to reinstall it with 'apt-get purge acpid'
  • Tried to fresh reinstall kernel
  • ACPI support checked as Enabled, also tried to enter it in config as = 1

If somebody already dealt with it :)
 
<markit> Nebraskka: just curiosity, if you open the console, do you see the shutdown process taking place and just not finisching, or seems that the signal is completely ignored?
<Nebraskka> markit, it's totally being ignored
<Nebraskka> i tried to log it somehow but didn't succeed
<Nebraskka> it's Ubuntu, i didn't find anything related to acpi in syslog or any acpi.log existance at all
<Nebraskka> the main difference - this VM was cloned from DigitalOcean, but others being installed from scratch and being cloned from each other. every each of them are shutdowning succesfully but DO clone
<markit> Nebraskka: no idea, try dmesg | grep acpi
<markit> Nebraskka: and if was installed before the rsync, I suggest you to remove + purge the package and reinstall again
<Nebraskka> markit, thanks :) tried to purge and reinstall it already, didn't worked out, but going to test dmesg way
<Nebraskka> thanks for idea
<Nebraskka> markit, hmmm http://upaste.me/5e7d17502321e3368


root@dev:~# dmesg | grep acpi
[ 0.000000] ACPI: LAPIC (acpi_id[0x00] lapic_id[0x00] enabled)
[ 0.000000] ACPI: LAPIC_NMI (acpi_id[0xff] dfl dfl lint[0x1])
[ 0.036534] acpiphp: ACPI Hot Plug PCI Controller Driver version: 0.5
[ 0.040088] acpi PNP0A03:00: _OSC: OS supports [ASPM ClockPM Segments MSI]
[ 0.040092] acpi PNP0A03:00: _OSC failed (AE_NOT_FOUND); disabling ASPM
[ 0.040137] acpi PNP0A03:00: fail to add MMCONFIG information, can't access extended PCI configuration space under this bridge.
[ 0.040326] acpiphp: Slot [3] registered
[ 0.040341] acpiphp: Slot [4] registered
[ 0.040354] acpiphp: Slot [5] registered
[ 0.040367] acpiphp: Slot [6] registered
[ 0.040380] acpiphp: Slot [7] registered
[ 0.040394] acpiphp: Slot [8] registered
[ 0.040413] acpiphp: Slot [9] registered
[ 0.040427] acpiphp: Slot [10] registered
[ 0.040441] acpiphp: Slot [11] registered
[ 0.040459] acpiphp: Slot [12] registered
[ 0.040474] acpiphp: Slot [13] registered
[ 0.040492] acpiphp: Slot [14] registered
[ 0.040506] acpiphp: Slot [15] registered
[ 0.040523] acpiphp: Slot [16] registered
[ 0.040537] acpiphp: Slot [17] registered
[ 0.040550] acpiphp: Slot [18] registered
[ 0.040564] acpiphp: Slot [19] registered
[ 0.040577] acpiphp: Slot [20] registered
[ 0.040590] acpiphp: Slot [21] registered
[ 0.040607] acpiphp: Slot [22] registered
[ 0.040621] acpiphp: Slot [23] registered
[ 0.040635] acpiphp: Slot [24] registered
[ 0.040648] acpiphp: Slot [25] registered
[ 0.040661] acpiphp: Slot [26] registered
[ 0.040674] acpiphp: Slot [27] registered
[ 0.040688] acpiphp: Slot [28] registered
[ 0.040710] acpiphp: Slot [29] registered
[ 0.040724] acpiphp: Slot [30] registered
[ 0.040737] acpiphp: Slot [31] registered


<Nebraskka> nothing happening here new after another shutdown
<markit> I've no idea indeed... try different kernel in the guest?
<Nebraskka> sorry for newbie question, any direction you could show to do that?
<Nebraskka> i tried to reinstall kernel from official packages
<Nebraskka> i agree that there is something could be in kernel that i copied
<speedlight> markit: trying to figuring out
<Nebraskka> but i didn't had any expertise with that earlier D:
<Nebraskka> only if it's not very time costly, could you point to somewhere related to this topic?
* Nebraskka touch markit
<markit> acpi_listen ?
<Nebraskka> oh, just a sec
<Nebraskka> root@dev:~# acpi_listen
<Nebraskka> button/power PBTN 00000080 00000000
<Nebraskka> button/power LNXPWRBN:00 00000080 00000003
<Nebraskka> yes, it receiving it
<Nebraskka> but not shutting down
<Nebraskka> hmmmmm
 
Worked around by commenting this section in /etc/acpi/powerbtn.sh:
# If logind is running, it already handles power button presses; desktop
# environments put inhibitors to logind if they want to handle the key
# themselves.
if pidof systemd-logind >/dev/null; then
exit 0
fi

and by adding this, to be sure, in /etc/systemd/logind.conf:
HandlePowerKey=ignore

Didn't managed to find out what was wrong with systemd-logind, but... oh well... :)