0

I have 7 pieces of Intel NUC8i5BEH2 (date of manufacture 04/2020). All of them with this configuration:

1xRAM Kingston 8GB 99U5700-014.A00G

SSD KINGSTON SA400S37/120G

BIOS updated with latest version 85.

Ubuntu 16.04 (same software configuration, I cloned SSD) with latest updates (as of 12/02/2020).

All of them are headless with HDMI Dummy plug (problem exist also with a monitor connected), then only connection are ethernet cable and USB3 where there is an Intel Realsense D415.

Since the very first moment (brand new hardware, clean install) I experienced random freezes, it can happen in a hour, or after a week.. but ALL of them experience this. They only thing you can do is hard reboot using the power button (don't reply to pings, keyboard inpunt..totally fronzen). The very strange thing is this randomness, when it happens, there is nothing strange in the logs. I also log memory usage and temperatures, and there is nothing strange also about that.

I know from the intel compatibility tool that Ubuntu 16.04 is not listed as validated by intel and the RAM is not in the validated list even if features are the same of the validated ones (while the SDD is).

I have never had this issues with the 7th generation of NUCs, basically with the same setup (we run a software that analyze the camera stream 12 hours/day).

There is also a scheduled shutdown every night and power on every morning.

Any jelp?

2
  • I have the same issue on Ubuntu 20.04 LTS (NUC10i5FNH). I have updated BIOS to FNCML357.0047.2020.1118.1629 thinking that might be the problem but no, randomly after 2 days the NUC froze again. I am not sure whether this is an Ubuntu issue or a hardware issue as I never had any other OS installed that I used. Did you find a solution?
    – Max101
    Commented Feb 25, 2021 at 12:04
  • I was not able to diagnose the exact problem, but Intel validate their NUC only for certain OS and hardware combination. So check if what you have is validated for that specific NUC model. One thing that I suspect is that could be related to the NIC, because I deployed on site my NUCs, where there is a high level network infrastructure and they haven't had any freeze in the last month. As a workaround I use a smart plug for each NUC and a script (in each NUC) which periodically pings all the others.. If someone doesn't reply, then the script power-cycle the associated smart plug.
    – rok
    Commented Feb 25, 2021 at 14:42

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .