Random crashes for seemingly no reason .

[DeterminedOpier]

This has happened now on identical brand new boards, new EMMCs, also on SD, identical 5a factory PSU, factory heatsink.

OS is Armbian, and it has happened on current desktop and server. This last time I had completed several successful reboots and thought maybe I just hadn't been pushing the plug in all the way. Then it happened again. Prior attempts have been on EMMC, directly burned with Etcher, and brand new Sandisk Ultra in package with bubble.

The behavior is while the board is running, white light flashing, system munching it just dies. Throws no errors. Just quits. Green light never goes out.

  I was tailing the emby log when it crashed this time.  No errors at all. Prior times it did not successfully reboot, but this one seems to be ok on reboot.

As I have been running Armbian for over 5 years on dozens of boards, this does not seem to be an Armbian problem, despite customer service at Pine claiming that this is a software issue. I have followed this exact same process on boards from Rpi2 and Opi PC to other RK3399 boards like NanoM4v2.  Not one has failed.

I am writing a book on NAS options (for people who don't actually know anything like me), so this is really relevant to the RP64. I personally bought two boards and NAS boxes and the card and everything.  Something is going on, and it is for sure hardware related.  This last crash was using an EMMC I got from Hardkernel years ago, and was unused. It seems like there was an issue with the 5 kernels at some point, but there are no notes about it in the current download page, and there are no links to legacy images.

 

Thanks as always. 

The pastebin is here:
https://paste.armbian.com/zikixujaho


Here is the syslog:

Apr 28 10:17:45 rockpro64 systemd-resolved[795]: Clock change detected. Flushing caches.
Apr 28 10:17:45 rockpro64 chronyd[1120]: System clock was stepped by 3624.542409 seconds
Apr 28 10:17:45 rockpro64 chronyd[1120]: System clock TAI offset set to 37 seconds
Apr 28 10:17:46 rockpro64 vnstatd[1094]: Info: Latest database update is no longer in the future (db: 2023-04-28 10:05:00 <= now: 2023-04-28 10:17:46), continuing.
Apr 28 10:17:48 rockpro64 systemd[1]: Starting system activity accounting tool...
Apr 28 10:17:48 rockpro64 systemd[1]: NetworkManager-dispatcher.service: Deactivated successfully.
Apr 28 10:17:48 rockpro64 systemd[1]: sysstat-collect.service: Deactivated successfully.

hardware log:

Time    CPU_cl0/CPU_cl1  load %cpu %sys %usr %nice %io %irq  Tcpu  C.St.
09:17:07  1416/1800 MHz  1.83  38%  21%  11%  0%  4%  0%  43.3 °C  0/5
09:17:08  1416/1800 MHz  1.83  39%  6%  28%  0%  4%  0%  42.8 °C  0/5
09:17:08  600/1800 MHz  1.83  39%  4%  31%  1%  1%  0%  43.3 °C  0/5
09:17:08  600/1800 MHz  1.83  40%  5%  31%  1%  2%  0%  42.8 °C  0/5
09:17:09  1416/1800 MHz  1.83  42%  9%  25%  0%  6%  0%  43.9 °C  0/5



Apr 28 10:17:48 rockpro64 systemd[1]: Finished system activity accounting tool.
Apr 28 10:18:01 rockpro64 systemd[1]: systemd-hostnamed.service: Deactivated successfully.
Apr 28 10:18:51 rockpro64 chronyd[1120]: Selected source 108.61.73.244 (0.ubuntu.pool.ntp.org)

[DeterminedOpier]

I burned the SD card and did a filesystem check of the EMMC, and looks fine.

 

root@rockpro64:~# fsck.ext4 /dev/mmcblk2p1
e2fsck 1.46.5 (30-Dec-2021)
armbi_root: clean, 62545/465120 files, 773907/1863680 blocks