TrueNAS SCALE RAIDZ1 pool causes kernel panic when one specific disk is attached

TrueNAS General
,
1

I’m running TrueNAS SCALE 25.10.4 (OpenZFS 2.3.4).

Pool configuration:

  • Pool name: CoffmanNAS

  • RAIDZ1

  • 3 × 16 TB Seagate Exos X16 (ST16000NM001G)

Issue:

  • System was running normally before I left for a week-long vacation.

  • I returned and found the system crashed.

  • On boot, TrueNAS kernel panics during pool import.

  • Panic message references:

    • zfs: adding segment overlapping with existing one

    • metaslab/space map loading

    • kernel panic during ZFS import

Troubleshooting performed:

  • Disconnected all data drives → TrueNAS boots normally.

  • Connected drive A only → boots.

  • Connected drive B only → boots.

  • Connected drives A+B → boots.

  • Adding the third drive consistently causes the kernel panic.

With the suspect drive disconnected:

zpool import shows:

pool: CoffmanNAS
id: 18434655501103277905
state: FAULTED
status: One or more devices contains corrupted data.

config:

CoffmanNAS FAULTED corrupted data
raidz1-0 DEGRADED
5566f80d-fb03-4519-b330-cebbaf319ba3 UNAVAIL
126e9f85-e436-4c88-bd75-57fcd23da53a ONLINE
8fba23d1-57d6-4ee8-8c63-c53a8639605a ONLINE

ZFS labels:

Disk 1 txg: 14173980
Disk 2 txg: 14173978
Disk 3 txg: 14173970

All three disks have readable ZFS labels via zdb -l.

SMART:

  • SMART passes on the suspect drive.

  • 0 reallocated sectors.

  • 0 pending sectors.

  • 0 offline uncorrectable sectors.

  • No logged SMART errors.

Import attempts:

  • zpool import -f -o readonly=on -o altroot=/mnt/recovery CoffmanNAS

    • Fails with I/O error.

  • zpool import -f -F -n CoffmanNAS

    • Returns immediately with no output.

  • Previous boot environments also panic when all 3 drives are connected.

Given that the suspect drive appears healthy and all pool labels are readable, is there a known OpenZFS recovery procedure for metaslab/space map corruption or an import method that may allow recovery of the pool or data?

IMG_9715C0A1-9816-4120-B23C-F4F3F715E9C3
IMG_9715C0A1-9816-4120-B23C-F4F3F715E9C31536×1152 407 KB

2

You can try importing it with -T to 14173970. But i would have a backup ready first.

3

The issue is that the system kernel panics during boot when the third drive is attached. I never reach a shell, so I can’t run zpool import -T 14173970. Is there a way to prevent automatic pool import at boot, or should I attempt the import from a Linux live environment?

4

I do have a backblaze backup of my important files

5

Try exporting the degraded pool with the 2 good drives first via GUI.

6

Did the drives pass a long smart test or just the short version?
You can try the import with the working two disks and rolling back transactions to 14173978. The pool will be degraded. If that is successful, I would back up all the data.
You may need to completely wipe the third drive and then try adding it back to the pool.

7

Thanks for your help all, I tried more steps and couldnt get it to work. I think im going to cut my losses and rebuild my nas using my backup. Thanks though!