3

We are experiencing a very particular and unique problem on our C# application which is running under .NET framework 4.7.2, in kiosk mode on Windows 10 Operating System that runs on a Lenovo ThinkStation P320 Tiny as Hardware (the same problem was also found on different hardware).

Environment Description:

the scenario includes an always running application, with automatic Day/Night Switch ON/OFF.
When the application is not in video call mode, it plays continuous loop of some videos/films.
The application runs on Windows precisely in kiosk mode at system start-up. The peripherals connected to the Mini-PC(Lenovo) are:

  • Primary monitor (Tablet touch)
  • Secondary monitor (TV 40")
  • Scanner
  • USB-Webcam
  • USB Microphone
  • USB LAN

The only possible user interaction with the system is performed in a controlled environment using a touchable tablet.

Problem Description :

in an apparently random manner we are experiencing a kind of machine freezes:
the primary and secondary monitors are turned on with a fixed image, network interface does not respond to ping command.
The operating system, in this time frame does not track any execution (no crash on the system event log, no dump, no active process running).
Even by connecting the mouse and keyboard to the Mini-PC it is not possible to regain control of the operating system.
There is no precise moment when this type of problem started to manifest itself as we started this project a few months ago.
The same problem was also encountered on different models of hardware different from the Lenovo ThinkStation P320 Tiny hardware used as target.

We have tried unsuccessfully to enable System Event Logs in the Lenovo BIOS Setup Utility but we have not encountered any errors.
We tried to force Dump generation from Windows but the operating system is not able to do this.
We have involved the Lenovo vendor (even if in reality the problem also occurs on different hardware) by performing diagnostic tools that can identify any hardware incompatibilities or intrinsic problems of our company.
Configuration: no problems were found.

In this situation the only possibility is to remove the Mini-PC power supply.
On reboot, by analyzing the application and the system logs, nothing is detected during the time frame in which the Mini-PC was blocked(frozen), no problem is detected in the period exactly before the freeze or after the restart.

Question:

Considering that neither from the application point of view nor from the operating system, it is not possible to get information about system diagnosis and problem analyzing, we would like to understand how to troubleshoot the problem, or enable low-level logs or dumps which can tell us something more about the status of the Mini-PC when it freezes or on the possible reasons which causes it.

systempuntoout
  • 1,895
  • 2
  • 12
  • 10
  • Dod you tried to replace the Hardware? – djdomi Jun 20 '19 at 16:35
  • Ad said, the same problem was also found on different hardware. – systempuntoout Jun 21 '19 at 10:57
  • 1
    Hi, is suggest to add into the PCI/PCIE Slots a debugging card to get the Error-Code in case of freeze, theyt might help you, if you dont get on a usual way in reference a device looks like https://www.amazon.de/ghfcffdghrdshdfh-4-Schichten-PCB-stabiles-Laptop-Motherboard-Diagnosewerkzeug-Fehlerbehebung-Diagnosekartentester/dp/B07RQQTZFK/ref=sr_1_5?__mk_de_DE=%C3%85M%C3%85%C5%BD%C3%95%C3%91&keywords=debug+card+pci-e&qid=1561440829&s=gateway&sr=8-5 @systempuntoout – djdomi Jun 25 '19 at 05:34
  • @djdomi the mini-pc boots without problem. Do you think POST card would be useful also for troubleshooting during the normal activity of the pc (when the c# program runs)? – systempuntoout Jun 25 '19 at 09:01
  • yes because it could tell you the reason for the hang – djdomi Jun 27 '19 at 20:35

1 Answers1

1

If the software cannot log the problem then it most definitely lies within a hardware fault. Usually, driver\software interaction problems cause kernal panic (BSOD).

Things to check:

  • Is the device overheating? (Ensure kiosk has good airflow)
  • Does the electrical source have any problems? (Missing ground/ground fault, voltage drops and spikes? Use a UPS to smooth out the power. Call electrician to fix ground faults.)
  • Faulty or cheap power supply. (This could happen on multiple devices that use the same PSU. Test the PSU.)
  • Large source of electro-magnetic interference nearby? (Example: MRI machines, Tesla coil)
  • BIOS bug. (Could exist in all devices with a certain firmware. Check for updates.)
HackSlash
  • 287
  • 4
  • 15