Pool is not healthy. So f...g what?

TrueNAS General
21

I would suggest removing the WD Purple for a better model of drive

Incidently thats a very similar build to my HairyNAS - its a solid build - apart from that disk. I used the 10Gb model of motherboard rather than your Quad 1Gb

22

This particular pool is only for the NVR VM and my Emby media. I don’t need a higher performance drive. They’ll all eventually be replaced with the same Purple while my other pool with all the important (and productive) stuff will stay on the WD Red Pro 10TB drives currently in there.

The old Synology RS820+ (slave NAS for backing up the TrueNAS important pool) still has the Seagate Ironwolf drives, as I like to go with different brands between appliances in case there are batch issues with manufacturers and whatnot.

I’m very happy with this motherboard. I was wanting to stick to an Atom CPU for low power consumption and it fit the bill with its capacity for 12 drives and an M.2 SSD.

I leave the grunt work to the HP DL20 hosting ESXi (eventually Proxmox when ESXi outlives its usefulness and Broadcom’s sales strategies). I’m not one to throw all my eggs in the one basket, as attractive as running VMs in TrueNAS might be.

23

Ah, now I understand.

No, ZFS will not tell you what files or metadata was affected by recoverable checksum, (or R/W), errors.

In general, ZFS should not care about 1 or more slower disk(s) in the mix. For RAID-Zx reading or writing that goes across the mixture, that will affect speed. Specifically, it will be limited to the slowest disk in the RAID-Zx stripe.

For me, I have a media server with 1 x 1TB mSATA SSD and 1 x 2TB 2.5" laptop HDD, (probably SMR), striped, (no redundancy), with ZFS. A big difference in speed, yet it works fine for media services use. Well, until I loose a block and have to restore a file. But I have good backups. Normally no checksum errors.

24

Hey, so I just got the error “Pool is not healty“
I saw here to type “sudo zpool status -v pool“

Which gave me:

pool: SSD
state: ONLINE
status: One or more devices has experienced an error resulting in data
corruption.  Applications may be affected.
action: Restore the file in question if possible.  Otherwise restore the
entire pool from backup.
see: 

scan: scrub repaired 0B in 00:05:34 with 0 errors on Sun Jan 25 00:05:35 2026
config:

    NAME                                      STATE     READ WRITE CKSUM
    SSD                                       ONLINE       0     0     0
      mirror-0                                ONLINE       1     0     0
        8ac3b09d-2533-4382-96a5-fbdc653abf46  ONLINE       1     0     0
        9372d4af-320d-4654-ac8f-50208ded3c03  ONLINE       1     0     0

errors: Permanent errors have been detected in the following files:

and then nothing… Sooooo? I got a read error but no files were affected?
Btw this pool is 2 NVMe SN700 and is ~1 year old. Should I be worried, do something about it?
I would guess a PCIe multiplexer failed for a few seconds which lead to this error?

25

I’d suggest to look at the smart data of the SSDs (smartctl -x <device>) to see if the SSDs are actually fine. To rule out a temporary error you might clear the errors and run another scrub.
Also this should probably be in it’s own thread.

2 Likes
26

Alright, thanks for the answer, a scrub fixed the error
I’ll know next time to create a seperate thread!
(not gonna create one now as it is fixed)

1 Like