Parity Memory Error when booting into TrueNas

TrueNAS General
1

This is going to be a long one, so I apologize in advance but I’m about to lose my mind.

TrueNas Mini X
ElectricEel 24.10.2
Intel Atom C3558 2.2GHz
16GB ECC RAM

When booting to my TrueNas drive (just one Samsung 870 SSD in bay #5 at the moment, no other drives attached), I’m getting an alert after logging into the TrueNas UI. Exact text (sorry, I’m not allowed to post pics) reads:

Sensor: ‘Sensor #0’ had an ‘Assertion Event’ (Software NMI ; OEM Event Data2 code = 00h ; OEM Event Data3 code = 28h)

There is a corresponding “Parity Memory Error” in the BIOS logs with Error Code: Smbios 0x03.

Now, some info that may be of interest:

  1. Memtest passed (granted, it only ran for ~4 hours), I also ran a SMART test on the drive and had zero errors. Tried installing on a spare Kingston I had lying around and it had the same “Parity” issues. So I think I can rule out the drive?

  2. When booting to a live USB or straight into BIOS, no “Parity Memory Error” is reported for that boot. Only when booting into the TrueNas install.

  3. The error doesn’t appear when rebooting after the initial install, only after every restart/shut-down thereafter.

  4. I’ve tried to figure out what “Sensor #0” in the alert could possibly be, but all the SEL documentation I can find doesn’t reference any “Sensor #0”.

  5. Similar for the Smbios error code. Except I think the docs I found were referring to a different model motherboard since there’s an event way back from August (before I had the Mini) with the code 0x16 which says “Log Area Reset” and the docs I found claimed 0x16 had something to do with North Bridge. They were also from 2015.

  6. I tried installing TrueNas on a standard (gaming) desktop using the same 870 drive. It installed fine and showed no errors or alerts, but it also doesn’t have the same ECC RAM or a server motherboard, so I’m not really sure what that proves.

  7. Not sure if relevant, but the clocks in the BIOS and TrueNas UI are both accurate. However the date for the alert in the UI is a couple minutes ahead (fast). Not sure how it’s getting the incorrect time.

  8. I verified the SHA-256 before installing, and installed following the instructions on the TrueNas site.

  9. I had this same exact issue on a previous Mini X, and after reinstalling I chose the option “load UEFI defaults” in the BIOS and the problem miraculously ceased. But I wasn’t keen on the idea of unexplained memory errors hanging over my head (especially given the use case) so I swapped it out for a new one. Well, I tried loading UEFI defaults this time, but to no avail.

If the problem were hardware it would seem odd to have the same exact error in the same situation (and only that situation) on 2 different systems. I’m leaning towards some sort of installation/BIOS setting/config needing to be tweaked, but I’m not getting anywhere.

Does it matter which bay the boot drive is in? PEBKAC? Perhaps a firmware update? Any help would be appreciated, I’m at a bit of a loss…

2

Okay, so I tried re-seating the RAM and reinstalling. That seems to have cleared up the issue, at least for now.

Still can’t guess why those other anomalies are the way they are (or were).

Could one be fairly confident moving forward with this build, or are there other tests I should run? (Besides SMART on the pool drives, that’s already planned).

3

If you browse a few other threads and take the Tutorial by the Bot, your forum trust level should go up and allow image posting.

TrueNAS-Bot
Type this in a new reply and send to bring up the tutorial, if you haven’t done it already.

@TrueNAS-Bot start tutorial