Cache drive issues - Direct IO failed

[LumpyCustard]

Approx every 2 weeks or so i run into an issue where my server goes into zombie mode, for lack of better word, where all my docker containers malfunction and VM's disappear.

 

I've gone through the logs (attached), and as far as i can see there are issues with devices SDD and SDE, which in my instance are two Samsung SSD cache drives connected directly to the motherboard (rather than the HBA that the rest of my array is connected to).

 

Both of the SSD's are less than a year old, and both were purchased separately at different times. The disappearance of my VM's and broken docker container i would assume is due to the fact that they're all hosted on my cache drive.

 

After rebooting my server, Unraid completes a full consistency check and always passes without issue, and none of my drives are marked as dirty.

 

I've just updated the BIOS on my motherboard (Gigabyte B450 Aorus Elite) in hopes that it will resolve, however i was hoping someone more knowledgeable could comb through my logs and advise what might be causing this.

 

 

Unraid Logs 1.txt devoraid-diagnostics-20210925-0011.zip

[JorgeB]

Problem with the onboard SATA controller:

 

Sep 25 00:02:12 devoraid kernel: ahci 0000:01:00.1: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xbfaf6000 flags=0x0000]

 

Quite common with some Ryzen boards, look for a BIOS update, if that doesn't help best bet it to use an add-on controller

[LumpyCustard]

Thank you, that is re-assuring. I was two BIOS updates behind.

 

I've updated as of now, will see what happens, but again, just wanted to make sure there was nothing else immediately obvious. I've noticed my server getting progressively less reliable (used to get months of uptime). 

 

F62

Update AMD AGESA ComboV2 1.2.0.3 B for Ryzen 5000 G-Series processors support

 

F63b

Improve memory compatibility