Well, thats not the email you want to get at the completion of a resilver of new larger drives… This is quite confusing since I have never had any errors in my array previously, and I run weekly scrubs. I recently added a L2Arc drive and set it to use metadata only, and run this pre-init script: ec…

[image] winnielinnie: Once a month is probably enough. You also want to lessen the chances of resilvers, SMART tests, and scrubs overlapping with each other. I do try and space out smart tests and scrubs etc so they don’t overlap. I will reconcisder my scheduels once I sort out the issue. …

Are those 7 drives with checksum errors on the same HBA? Not related the your issue, but worth mentioning: [image] ligistx: weekly scrubs Once a month is probably enough. You also want to lessen the chances of resilvers, SMART tests, and scrubs overlapping with each other.

[image] ligistx: Unless you think this is a terrible idea, I think my plan is do a scrub with fans at 100%. That sounds wise. Do the usual, on top of bringing down the temps. Check all connections, on both ends, and the HBA’s seating itself. When you have multiple drives on the same HBA simu…

Fans at 100, everything is cool to the touch when I check… You anticipate it can clear the metadata error? What I don’t fully understand is how it would do that, unless the error is not real? I always fear the long debunked “scrub of death” in such a situation… but scrub is 10% done, so I suppose w…

[image] ligistx: You anticipate it can clear the metadata error? It’s possible that because 7 disks in a RAIDZ2 vdev “errored” simultaneously, there was not enough parity available to properly read the ZFS metadata at that point in time. This gets registered as a “permanent error”. If you co…

Ah, that does make sense. Thanks for the explainer! I did a zpool clear prior to the scrub, but I suppose if it doesn’t clear up the metadata “permanant error”, there is no way to clear that fault? Seeing the pool as unhealthy for the rest of my life is probably going to kill me ;(

I would try it again after a full successful scrub that returns 0 errors after completion.

Woke up to a fully healthy pool. No errors reported at all, and the permanent error was no longer being reported. Thanks for the advice!

Sounds like it was the HBA after all. Keep those temps low. :+1:

And one Smiling Mask Award ! :clown_face:

I was definitely sweating there for a bit. I have had issues in the past due to bad cables or poor plug connections, but usually that’s 1 maybe 2 drives. I had never had a metadata issue which is quite scary. But agreed with all the above, the fact 7 drives all at once had seemingly related issues……

Permanent errors have been detected in the following files: <metadata>:<0x1b>

TrueNAS General

etorix September 9, 2025, 3:31pm 11

And one Smiling Mask Award!

Topic		Replies	Views
ZFS read/write errors during scrub TrueNAS General SCALE , ZFS	21	1675	January 29, 2025
Write errors during pool extend - what do I do now? TrueNAS General SCALE , ZFS , Replication	24	297	March 28, 2025
Weird CHKSUM errors after drive replacement TrueNAS General SCALE , Hardware , ZFS	48	314	August 27, 2025
Resilvering is Cooking My Drive TrueNAS General CORE , Hardware , ZFS	48	964	January 11, 2026
My pool appears to be in serious trouble TrueNAS General SCALE , Hardware , ZFS	19	280	October 1, 2025

Permanent errors have been detected in the following files: <metadata>:<0x1b>

Related topics