UNRAID Forcing Graceful Shutdown on HPE Proliant Server

[bluiwulf]

I've been searching all over for a potential fix to a sudden problem I am having w/ UNRAID 6.10.3.  It keeps reporting a sudden power button press and graceful shutdown.  The power button isn't getting pressed and I've done extensive diagnostic testing and on the server and cannot find a problem.  The only reported hardware issues are w/ the Smart Array Battery, which I am getting a replacement for.

 

I also ran Memtest and even went as far as to install TrueNAS and verified that the server does not initiate a shutdown in any other situation except being booted into UNRAID.  I reinstalled the OS on the flash drive and attempted to use two additional flash drives, but the issue persists.

 

I would appreciate any other suggestions for troubleshooting this issue.  It's a HPE Proliant DL380 Gen9 w/ 2x Xeon processors and 64GB RAM. 

[JorgeB]

First thing to try is booting in safe mode to rule out any plugin, you can also disconnect the power button just to make sure there's not a problem there.

[bluiwulf]

I did attempt safe mode a few times and disabling the front panel buttons and USB.  It didn't make any difference.

[JorgeB]

Boot in safe mode, enable the syslog server, and post that after it shuts down.

[bluiwulf]

Here is the syslog and the diagnostics.  I'm going to test each of the memory sticks next just to make sure they are each good.

biest-diagnostics-20220708-1850.zip syslog.log biest-diagnostics-20220708-1846.zip

[JorgeB]

I assume it shutdown between these?

 

Jul  8 18:06:44 Biest rc.inet1: ip link set lo down
Jul  8 18:45:06 Biest kernel: microcode: microcode updated early to revision 0x49, date = 2021-08-11

 

If yes, not shutdown command, so possibly a hardware issue.

 

 

[bluiwulf]

18 minutes ago, JorgeB said:

I assume it shutdown between these?

 

Jul  8 18:06:44 Biest rc.inet1: ip link set lo down
Jul  8 18:45:06 Biest kernel: microcode: microcode updated early to revision 0x49, date = 2021-08-11

 

If yes, not shutdown command, so possibly a hardware issue.

 

 

 

Actually no, I keep a secondary monitor on connected to monitor for it and this is when it starts the graceful shutdown:

 

Jul  8 18:45:35 Biest webGUI: Successful login user root from 10.0.0.3

Jul  8 18:50:52 Biest elogind-daemon[1977]: Power key pressed.

Jul  8 18:50:52 Biest elogind-daemon[1977]: Powering Off...

 

[JorgeB]

Something is missing, a shutdown either by using the GUI or a quick press of the power button, should always be logged like this:

 

Jul  9 08:53:04 Test2 kernel: md: sync done. time=2860sec
Jul  9 08:53:04 Test2 kernel: md: recovery thread: exit status: 0
Jul  9 08:53:31 Test2 shutdown[48304]: shutting down for system halt
Jul  9 08:53:31 Test2 init: Switching to runlevel: 0
Jul  9 08:53:31 Test2 init: Trying to re-exec init

 

Please test that yourself by actually pressing briefly the power button.

[JorgeB]

Ignore the above, missed the times:

 

Jul  8 18:50:52 Biest elogind-daemon[1977]: Power key pressed.
Jul  8 18:50:52 Biest elogind-daemon[1977]: Powering Off...
Jul  8 18:50:52 Biest elogind-daemon[1977]: System is powering down..
Jul  8 18:50:52 Biest elogind-daemon[1977]: Suspend key pressed.
Jul  8 18:50:52 Biest shutdown[5530]: shutting down for system halt
Jul  8 18:50:52 Biest init: Switching to runlevel: 0

 

Is this elogin something you installed? Can you try without it.

[JorgeB]

Possibly related to this?

https://forums.unraid.net/topic/3531-looking-for-better-ideas-how-how-to-sleepsuspend-my-unraid-box/?do=findComment&comment=1128998

 

[bluiwulf]

I attempted this and it just forces a suspend state that locks up the entire server.

[bluiwulf]

Well I was able to use that link to provide a work around.  I changed the /boot/config/go file w/ the following lines to software disable the power switch:

 

# Change power button handling and restart elogind to reload the edited config

/usr/bin/sed -i -e 's/#HandlePowerKey=poweroff/HandlePowerKey=ignore/g' /etc/elogind/logind.conf

/etc/rc.d/rc.elogind restart

chmod -x /etc/acpi/acpi_handler.sh