Seemingly random, seemingly traceless crashes: Needing help in diagnosis

User Support Hardware Issues
1

Hi. For a while now, I have had seemingly spontaneous crashes while using Qubes, across releases 4.2 and 4.3, and across several Xen and kernel patch versions, and a few minor versions.

  • It does not seem to be related to overheating, nor power loss
  • It happens both when the device is in active use and when it isn’t, and when it is charging, and when it is running off its battery
  • It does not happen after a consistent amount of uptime, sometimes occuring within half an hour of a previous crash, sometimes taking several days to occur again, and everywhere in between
  • The journal does not show anything, just normal system activity, then the next boot header
  • The same happens in /var/log/xen/console/hypervisor.log

The screen simply goes blank, before the power light briefly turns off, before the system turns on again, showing no signs of an error. I don’t think (and hope it isn’t) a hardware issue is at fault, since the hardware (a 7040 series Framework 13) is fairly new, and has been well taken care of.

Any help in diagnosing and fixing this issue is greatly appreciated. :slight_smile:

2

I’ve had similar issues on different hardware and it turned out to be related to regular power surges/brownouts in my area.

3

That’s interesting, but it wouldn’t explain why these crashes still occur when on battery power alone.

4

Right, reading comprehension…

Do you have a way of testing whether the same happens on other Linux distros? That would narrow down whether the problem is Qubes or something else.

5

I could try using a live distribution. What would you suggest? Tails?

6

Use whatever your threat model permits. Tails should be good.

7

I had similar crashes when I had a bad battery. It ran fine on mains only, but failed any time the battery was present. I decided it was basically brownouts caused by short circuits somewhere in the battery.

Lots of other possible causes - damaged circuitboard track, dry solder joint, bad contact in cpu socket or other connection… and many others.