Server keeps turning itself off

[TryRebooting.Tech Unraid]

My server has kept turning itself off over the last several days.  I recently replaced a drive that was going bad, and it's been stable for a handful of days with the new drive while it did the data rebuild.

I closed up the machine and within a day or two it started to power off randomly.  After hitting the power button, the machine comes back up, boots properly, and works as expected.

Attached are my log files.  I'm seeing this line but not finding anything specifically after a Google search:

Jan 11 16:34:15 Unraid kernel: ACPI Warning: \_PR.C003._PSS: SubPackage[0,1] - suspicious power dissipation values (20220331/nsrepair2-684)

syslog

[trurl]

Do you have an UPS?

 

Even if it crashes it doesn't power down. And you are having unclean shutdowns. Are you sure it is powered down and not just crashed? Are you sure you aren't losing power to your server? Somebody (or something) messing with the power button when you aren't looking?

[TryRebooting.Tech Unraid]

Thanks for the reply.  I have a UPS but it's not connected via USB, just something it's plugged into.  

When I come into the room it's totally powered off, or at least it appears that way.  Fans not spinning.  RGB lightings from the motherboard is off.

I'm not losing power to the server that I know of, no breakers tripped or anything.  No one messing with the server either, 100%.

I can move the plug to a different outlet and see what happens.  Can you tell me/show me what in the log you're looking at?  I have to admit I get the gist of the lines but not the bulk of it. 

[TryRebooting.Tech Unraid]

Before I moved the machine to a different outlet and bypass the UPS, I found it was powered off again.  Here is a copy of that log file in case it is helpful.  

Meanwhile, the machine is moved to a different power outlet and without the UPS.  If it's going to power off again I suspect it'll happen in 2ish hours or so.

syslog (1)

[trurl]

Instead of syslog we always prefer diagnostics. Diagnostics has current syslog and many other things about your hardware and configuration that helps make sense of your situation. I seldom look at syslog until I have already looked at other things in diagnostics to give syslog some context. For some kinds of problems I don't even look at syslog at all.

 

 

[TryRebooting.Tech Unraid]

Sounds good.  I'll try to run that, although I just found my server is off again.  So the shutdowns are still happening and appear to be more frequent.  I'll try to boot and run that.  Thanks!

[trurl]

Just now, TryRebooting.Tech Unraid said:

just found my server is off again.

Something wrong with power supply or power cables or connections to motherboard.

[trurl]

Does it power off if you don't boot anything and just let it sit at BIOS screen?

[TryRebooting.Tech Unraid]

I could pull the Unraid USB to test but I run it headless so I wouldn't be able to tell what the boot was doing.  But I would be able to tell if it was still on or not.  I can try that next.  I was able to get it to boot again and I ran the diagnostic. Hopefully this can get us some specific info. 

I'm going to give a visual inspection to the power cords, see if anything is loose or whatnot.

 

unraid-diagnostics-20230111-2010.zip

Edited January 12, 2023 by TryRebooting.Tech Unraid

[TryRebooting.Tech Unraid]

The plot thickens.  Okay, so the cables all seemed fine.  I unplugged USB and booted.  I went back into the room to see how the machine was doing.

I noticed it seemed to be mid turn-off right when I was in there looking.  I think the CPU fan might have died.  Doesn't seem to be spinning, and the CPU heatsink is screaming hot.  Might be the cooler died.  I'm going to let it cool down, then boot it back up to briefly test if that fan spins.

Would explain why it's powering down faster now, since there is more residual heat.

If you see anything in the diagnostic that supports this I'd be all ears.  Thanks for your help either way, nice to have someone to talk to about this.

[TryRebooting.Tech Unraid]

I think that's it.  CPU cooler died (hopefully) or power to the cooler is jacked.  Fan spins for a bit, stops, spins, stops, spins, stops.  Then I shut it down.