Raid Array Problems with NVMe Drives

Insight & Feedback
1

Good day. Is there some problem with using NVMe drives as the primary storage medium within ZimaOS+? I have configured five NVME drives into both a RAID1 and RAID0 arrays always working from within the OS WebUI (except when wiping the drives in between raid arrays). This system does not have any spinning rust drives. (Thankfully I got everything before the prices went Artificially Insane.)

In both raid configurations I have encountered multiple different errors after copying some large (legally obtained) movie files onto the arrays during multiple copying sessions. The error messages start off by showing that the arrays are in a “Protected” state. They always wind up in a read-only state. This happens after copying over only a few hundred gigabytes of files. This makes the arrays useless for my purposes.

All five SSDs are all almost brand-new. All drives show up as being “Healthy”. The “Remaining Life” statistic for all five drives shows 100% (0% Used). During the copying process the ZimaOS+ Dashboard shows that the CPU (N150) loading is at 40% with about 10% of the system ram being used (16GB total). Nothing in the system is over 50 degrees C.

While poking around I have found a great many raid errors (dmesg | grep -Ei “md0”) and many Btrfs errors (dmesg | grep -Ei “btrfs”).

Sadly, it’s looking like this is a ZimaOS+ problem. I have installed the 800-pound gorilla of free NAS software onto this very same system without any other modifications. I have copied the same hundreds of gigabytes (and more) onto the system over multiple different sessions without any hint of a problem.

For my purposes I would much prefer to use ZimaOS+ rather than something more complicated and cumbersome. Any help would be appreciated. Please let me know what information I could provide that might help diagnose these problems. Please include the actual commands as I’m not very knowledgeable at this time.

Thank you for any insight you could provide.

1 Like
2

While this might be a dead thread, I did want to update it with my latest findings.

After reinstalling TrueNAS for the second time I put the raid array through a more strenuous test by copying > 500GB onto it over two copying sessions. The first copy went flawlessly. After the second, four of the five drives had been exported from the raid array, and the fifth one was unreachable.

After some digging, I found that there was a problem with one of the NVMe sockets on the motherboard. I swapped two of the other drives into that socket with very similar results. Funny enough, after rebooting the machine after the the array had failed with the third drive in the slot, the problems cleared up and the array is now working properly. Unfortunately, I have many more movie files to copy onto the machine.

Long story short, it’s time to send the motherboard back to the manufacturer for repairs or replacement (I guess I’ll see how well they honor their 3-year warranty).

1 Like
3

Thanks for the update. It sounds like the issue is likely related to the faulty NVMe slot rather than ZimaOS.

Glad you were able to identify the cause, and thanks for sharing the findings with the community. Hope the warranty replacement goes smoothly!