Hello all, I’m a new user and getting a warning of disks with “non-unique” serial number. However, when I open Disks in the web interface it shows different serial numbers.
I’m using a MeMini with 4 Crucial 4TB NVMe M.2 SSDs. The OS is on the drive that came with the mini PC, not any of the ones I added.
I’d like to add them to the pool but am obviously concerned about data loss, given the warning.
Weird, this usually happens when people use usb drives, but looking at pictures of the ME Mini, the drive at least seem to connect directly to the motherboard.
No chance that you’re running virtualized or with raid enabled in bios, or anything else that can prevent direct access to the drives?
I wonder if these NVMe are knock offs, and not real Crucial NVMes?
Or perhaps they have been “refreshed”, (aka a scammer took warranty returns, and made them look “new” again).
Not sure how to tell in either case. But, sometimes physical examination could give a hint.
No sign of that in the BIOS. And I didn’t get this error when I was trying out Open Media Vault… Odd indeed.
I don’t think they’re knock-offs but I’m realizing now that while I thought they were coming from Amazon, they weren’t – rather a reseller. Far too late to get a refund. I’m going to test them further. (Open Media Vault didn’t have a problem with them when I was testing it.)
Update:
I tested these drives with Crucial’s own software (“Crucial Storage Executive Tool”) and it saw nothing problematic with the drives. So I’m fairly comfortable that they’re legit after all.
Which leaves me with the dilemma of whether to just go ahead and use them in the TrueNAS setup or use another NAS system…
If wouldn’t be surprised if this is another peculiar interaction with the eMMC storage and certain partitions that get created on those. There’s a similar report over at the beelink forum.
I have seen similar situations when truenas is implemented over some kind of storage solutions - like Intel’s, or RAD controllers.
it hasn’t been a problem for me.
That makes the most sense from this situation
Same problem, it gives me the same code 12345. I'm wondering if anyone can help me online with a virtual connection. If they can connect to my computer to fix the problem, I have two Western Digital Red Plus 4 terabyte drives.
Going to use some other server software, I guess. GIven that nothing else I’ve tested (Open Media Vault, UnRaid, Jellyfin) shows any issue of this kind – and that testing with the drive manufacturer’s own tool showed discrete serial numbers – the issue seems to be with TrueNAS. Regrettable, because I’d settled on it given its feature set and open nature. Still not sure which of the alternatives I’ll use…
Another user reports a similar issue about (spurious) duplicate serial numbers in the boot pool while attempting to create a data pool.
I suggest you file a bug report over this.
I have experienced this same issue after installing one of the recent 25.10.x (Goldeye) versions on my home lab server.
I verified serial numbers were reporting uniquely in software. Truenas reads these serials slightly different than other OSes.
I had previously installed 25.04.2.5 (Fangtooth) on this same hardware and was able to create a VDEV with the NVME drives without issue.
So, whatever change occurred, it happened between those releases. I have reason to believe it is related to the logic associated with the serial number checks as my affected drives have serial numbers like these:
”ACE4_2E00_4502_0123_2EE4_AC00_0000_0001.”
”ACE4_2E00_4502_0456_2EE4_AC00_0000_0001.”
Where only the “_XXXX_” number in the middle of the, terribly long, serial number is unique. Also these serial numbers have ”.” which may trip up the check logic if it were updated to use regex or some other similar comparison mechanisms.
For my home lab purposes, I went back to 25.04.2.5 for now.
I do not typically post on forums such as these, so please forgive any etiquette I might have missed. This appears to be a recent issue since 25.10.x and thought sharing my experience would help.
I get this warning only on the two eMMC boot partitions. I am running on the latest BIOS from Beelink, so that might be a thought if you haven’t updated.
I am having the same problem on 25.10.1 adsf, but I require Goldeye because of the recent patches that were applied from upstream openZFS (long waited fixes to encryption).
I agree the problem seems to happen when the drives have very similar serial numbers… I am having problems with these two pairs of drives:
Model Number: Netac NVMe SSD 1TB
Serial Number: AA202510151T65004157
Model Number: Netac NVMe SSD 1TB
Serial Number: AA202510151T65004097
Model Number: INTEL MEMPEK1J032GA
Serial Number: PHBT91350569032P
Model Number: INTEL MEMPEK1J032GA
Serial Number: PHBT913503TD032P
Have got the same problem. TrueNAS UI reports unique serials numbers. After first try I not even get the option toi ignore.
Any chance for a fix?
Unfortunately I had to disassemble the Truenas Home-lab configuration I was using and had to utilize the SSDs for another project. So I am unable to reproduce the issue at this time.
I am no expert in Truenas but I believe it is related to the logic associated with the serial number checks. Possibly check the logs when you encounter this error again and hopefully it will identify the source of the log entry.
Based on my limited search, I think issue may reside in the pool creation code where it would be checking if serials are identical. I am no expert in Python, but it might be associated with this..
middleware/src/middlewared/middlewared/plugins/pool_/pool.py at e1af25fc517f4fe8209881a42a7903e88219efcc · truenas/middleware · GitHub -197.
If I had more free time I would try dig in and identify the issue myself, then send in a report.