Resolved Mac Mini 2018 Kernel panic "x86 CPU CATERR detected" when OWC Thunderbay 4 connected

[djc6]

For the past week, my 2018 Mac Mini has been having a kernel panic "x86 CPU CATERR detected" a few times a day. I noticed it stops when I disconnect my nine year old OWC Thunderbay 4 (Thunderbolt 2 edition - purchased 2/9/2015). This is what I have tried so far:

1. Different thunderbolt ports on the mac mini so that it would use Thunderbolt controller 0 vs Thunderbolt controller 1
2. Different thunderbolt 2 cable
3. Different thunderbolt 2 to 3 adapter
4. Different port on the back of the Thunderbay 4

None of these have helped. Any other suggestions short of trying another disk enclosure? They are like $500 so that would be an expensive experiment.

I'm trying to figure out if the computer or enclosure is bad. Running Diagnostics (holding down "D" key on startup) says my mac mini is fine.

I might have access to my father's m2 mac mini for testing, to see if its unhappy with the Thunderbay 4.

 

[joevt]

I suppose the only thing you haven't tried (besides a different macOS) is putting a Thunderbolt device between the Mac mini 2018 and the OWC Thunderbay 4.

 

[HDFan]

What does OWC support say?

 

[djc6]

Thanks for the ideas. I'm going to try putting my Caldigit TS3 hub from my laptop between the drive and computer, see if that changes anything. I also wrote OWC to see if they know of any ideas.

I'm going to try seeing if it has anything to do with which drive bays are populated - I pulled two drives (a Disk Utility RAID1 array) to see if that changes anything.

I routinely use a XPOWER Airrow Pro electric duster on the unit to keep it cooling properly. Today when I dusted I looked inside with a flashlight and I was disappointed when I opened it to see so much flux residue left over from the manufacturing process - not sure if that can be a source of problems. At the bottom of the board you can see a lot of flux residue around where the two thunderbolt connectors are on the back of the unit, but there is dried flux residue around any through-hole components.

 

[djc6]

Another fun variable! I have an eight year old (Aug 2016) CyberPower CP1500PFCLCD UPS that both the mac mini and the Thunderbay 4 are plugged into and nothing else. Today when my daughter used the brother laser printer, it caused the voltage to sag just a little so that the UPS kicked on, except its apparently not working so the mac mini and thunderbay powered off suddenly. So I wonder if the flakey UPS providing power to 2018 mac mini and thunderbay have anything to do with this.

 

[djc6]

So every day I've been putting one of the drives back into the Thunderbay 4. When I woke up, bays A, C, D were populated. I put the drive for bay B back in and its crashed three times so far this morning when I haven't seen any crashes since Monday when I started pulling drives.

Curiously, the drive in bay B is my time machine backup drive, and I can't mount it anymore. Disk Utility when I run "First Aid" on the drive itself quickly (in seconds) returns that its fine. But when I try and run "First Aid" on the HFS+ volume itself, I get an error right away "First Aid process has failed". Sometimes it says error #8 and sometimes error #-69845 - possible drive is bad and not just the filesystem? I might try reformatting it. The Time Machine drive itself is almost six years old, purchased June 2018 so almost 6 years old. Its lead a hard life compared to my other drives, with time machine running hourly - thats like 50,000+ backups!! I still have another backup of my system on Backblaze.

I also cannot read or run First Aid on this HFS+ volume when its inserted into another bay, so I don't suspect the bay is bad since the issue follows the drive to other bays.

I emailed OWC on Sunday, received an automated response acknowledging receipt of my email, no human response so far.

 

[Fishrrman]

OP:
What you describe in reply 6 above sounds like "a drive problem", and not "a bay problem".

I would take the backup drive and put it into either:
- a standalone external drive enclosure
or
- a USB3.1 gen2 SATA docking station
and then
... see how the drive behaves "on its own".

If it still behaves badly, then it's time to replace it.
Time to start fresh with a new drive AND a new backup.

 

[djc6]

I tried the suspect drive on another computer (m2 16" macbook pro) with another enclosure, and I can't format the drive. It starts to, but then you hear the drive spin up/spin down like it had power abruptly cut to it. Disk Utility would just hang and require Force Quit. Made multiple attempts to try and partition/format the drive.

I've not had any kernel panics on my 2018 since pulling this bad drive, the issue is solved. Hopefully this helps others - many posts regarding "x86 CPU CATERR detected" have a default response that you a new logic board. Try the computer without any accessories first and see if you can whittle it down to a problem device.