Seems very serious . . .

[Glorious1]

For weeks I've been trying to diagnose a problem with my Kodi installation, which streams media files from my FreeNAS via NFS. Lots of random freezing. Kodi logs are inscrutable.

The last time it happened, I had some kind of watchdog event on FreeNAS. It was basically frozen. As I recall, it did not respond to a software shutdown, but I was able to do an immediate shutdown via IMPI.

Since then, FreeNAS either
a) won't start at all; the machine is stuck at the SuperMicro splash screen, with "System initializing . . ." in lower left corner, or
b) eventually gets past that splash screen, but the IPMI is apparently messed up: my fan script which calls IPMI sdr, gets a response: "Could not open device at /dev/ipmi0 or /dev/ipmi/0 or /dev/ipmidev/0: No such file or directory".

When this happens, the IPMI gui becomes partially nonresponsive. Resetting the IPMI or pulling the plug does not fix it. One time a got an email about a nonrecoverable voltage problem, but that only happened once and may have been spurious with everything else going on.

I'm scared and would appreciate some suggestions on diagnosing this.

 

[m0nkey_]

What model SuperMicro board?

 

[Glorious1]

What model SuperMicro board?

Here is the info from my signature:

SuperMicro A1SRi-2758F mini-ITX with integrated
CPU: Intel Atom C2758 CPU, 8-core 2.4 GHz
HBA/Expander: LSI 9240-8i PCIe board flashed to IT mode, provides 8 SATA ports
RAM: Crucial 2x 8GB SO-DIMM 204-pin Unbuff. ECC DDR3-1600 1.35V
Drives: 3x WD Red NAS 3 TB, 4x WD Green 3 TB
Power Supply: SeaSonic Platinum Series SS-400FL2 Fanless, Active PFC
Case: Lian Li PC-Q26 with 5x 120-mm fans and room for 10 3.5" drives
FreeNAS: FreeNAS-11, four disks in RaidZ1 (remaining disks for backup)
Boot Device: 2x mirrored SanDisk 8 GB Cruzer Fit

 

[Stux]

Has your system had the Avoton repair made?

 

[Glorious1]

No, never heard of it. ??

Edit - I did some research on the Avoton problem. Seems like it fails usually 18 months to 3 years. This is just short of 3 years. It's out of warranty, but I'll see if Supermicro will replace it.

https://www.servethehome.com/intel-atom-c2000-series-bug-quiet/
https://www.theregister.co.uk/2017/02/07/intel_atom_failures_go_back_18_months/

 

[Stux]

Many suppliers extended warranties on Avoton systems

 

[Glorious1]

Short story is, Supermicro looked it over, found nothing wrong, and I've still got problems.

They agreed to accept an RMA even though it was out of warranty. That's nice. They looked at it, updated the bios, did some other stuff I don't understand, and said they couldn't reproduce the problem. They returned it.

Since all my BMC-IPMI settings were wiped, it had to wait till I had time to play with it. I was able to boot reliably (it was uncommon before), but the fans are no longer behaving. I reset the fan thresholds so they are outside the range of the fans.

The problem is, the fans always stay at 100% duty cycle. If I restart with no script running, everything cold, they run 100%. It doesn't matter if fan mode is Standard (the lowest mode for my board).

When I run the script, and it commands lower duty, I can hear fans momentarily become quiet, then in less than a second, back to 100%. Any advice would be greatly appreciated.

Here is what I submitted for RMA:

I had a nonrecoverable watchdog event, then the board wouldn't start. It gets stuck right at the beginning, with the Supermicro splash screen and the "System initializing . . ." message. Occasionally it will boot but the IPMI is not functioning correctly, and resetting IPMI doesn't fix it. I suspect this is the known Avoton hardware issue, and the time is right, a bit less than 3 years. Can you replace it?

I was able to get the repair report:

REPORTED PROBLEM NOT FOUND. ECO REWORK DONE BY RR14. VERIFY ECO. UPDATE BIOS. CLEAR CMOS BATTERY. TEST ALL DIMM SLOTS.BOOT INTO WINDOWS OS. VERIFY ME.TEST PASSED.

EDIT: After resetting the IPMI a few more times and rebooting, the fans are under control again. But I don't trust this motherboard any more.