Hey guys,

i have an issue with my ssd cache Pool. I set it up as raidz1 from 4x 1TB SSD. But one ssd often gets offline. The Overview shows in STATE the REMOVED for SSD. In the last Column of the state table it states "(trimming)". Have i misconfigured something? I know ssds need trimming for their health, but i think the ssd never exits the state when it enters it.

Additionally one disk runs in FAULTED state because of corrupt data. But this is not the problem actually as there is for now no important data on the pool.

So what do i have to do about the trimming?
Pool state:

When i restart the server the pool is back online but the same disk gets removed after some time again

Thanks for any help.

There are issues with two devices, not just one; please post the diagnostics.

Post diagnostics, but sounds like that drive is disconnecting from the system.

2 minutes ago, JorgeB said:

There are issues with two devices, not just one; please post the diagnostics.

As stated i have seen the failure with the fourth disk (sdj1), but my point in this thread is that the disk keeps getting marked as "removed" due to "trimming."

However, here are the diagnostics

unraid-diagnostics-20260516-1355.zip

5 minutes ago, Kilrah said:

Post diagnostics, but sounds like that drive is disconnecting from the system.

In the "Main" view the disk is online and i can start smart checks.
In lsblk the disk is also shown, but without any partition:
sdh 8:112 0 931.5G 0 disk

└─sdh1 8:113 0 931.5G 0 part

sdi 8:128 0 931.5G 0 disk

└─sdi1 8:129 0 931.5G 0 part

sdj 8:144 0 931.5G 0 disk

└─sdj1 8:145 0 931.5G 0 part

sdk 8:160 0 931.5G 0 disk

There's a problem with the pool cfg file, try this first:

on main click on the first device for that pool and then "remove pool"

back on main, create a new pool with the same name and number of slots

assign the pool devices, leave the filesystem set to auto

start the array to import the pool and post new diags.

11 hours ago, JorgeB said:

There's a problem with the pool cfg file, try this first:

on main click on the first device for that pool and then "remove pool"

back on main, create a new pool with the same name and number of slots

assign the pool devices, leave the filesystem set to auto

start the array to import the pool and post new diags.

Here we go. I had to hard shutdown my server. As the crashed pool was my cache Pool, i had multiple Zombie processes of stuck docker container waiting for io...

unraid-diagnostics-20260517-2114.zip

I don't see an attempt to import the pool, did you leave the filesystem set to auto as mentioned? Also post the output from zpool import

1 hour ago, JorgeB said:

I don't see an attempt to import the pool, did you leave the filesystem set to auto as mentioned? Also post the output from zpool import

yes i left it in auto. The Main view suggests formatting the disks.
zpool import:

root@Unraid:~# zpool import
  pool: cache
    id: 12246082903798894337
 state: UNAVAIL
status: One or more devices contains corrupted data.
action: The pool cannot be imported due to damaged devices or data.
   see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-5E
config:

        cache       UNAVAIL  insufficient replicas
          raidz1-0  UNAVAIL  insufficient replicas
            sdi1    ONLINE
            sdh1    ONLINE
            sdk     UNAVAIL  invalid label
            sdj1    FAULTED  corrupted data

Pool has two bad drives at the moment; since it is raidz1, it can only recover from one, so you will likely need to recreate it.

See if this works:

zpool import -o readonly=on -F cache

Also post output from

fdisk -l /dev/sdk

5 minutes ago, JorgeB said:

Pool has two bad drives at the moment; since it is raidz1, it can only recover from one, so you will likely need to recreate it.

See if this works:

zpool import -o readonly=on -F cache

Also post output from

fdisk -l /dev/sdk

root@Unraid:~# zpool import -o readonly=on -F cache
cannot import 'cache': one or more devices is currently unavailable
root@Unraid:~# fdisk -l /dev/sdk
Disk /dev/sdk: 931.51 GiB, 1000204886016 bytes, 1953525168 sectors
Disk model: SanDisk SSD PLUS
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 16776704 bytes

That drive is missing the partition type

sfdisk /dev/sdkthen type 2048 and hit enter and post the result (don't type anything else for now)

4 minutes ago, JorgeB said:

That drive is missing the partition type

sfdisk /dev/sdkthen type 2048 and hit enter and post the result (don't type anything else for now)

It instantly gives me a warning about zfs tag:

root@Unraid:~# sfdisk /dev/sdk

Welcome to sfdisk (util-linux 2.42).
Changes will remain in memory only, until you decide to write them.
Be careful before using the write command.

Checking that no-one is using this disk right now ... OK

The device contains 'zfs_member' signature and it may be removed by a write command. See sfdisk(8) man page and --wipe option for more details.

Disk /dev/sdk: 931.51 GiB, 1000204886016 bytes, 1953525168 sectors
Disk model: SanDisk SSD PLUS
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 16776704 bytes

sfdisk is going to create a new 'dos' disk label.
Use 'label: <name>' before you define a first partition
to override the default.

Type 'help' to get more information.

>>> 2048
Created a new DOS (MBR) disklabel with disk identifier 0x61c46fa5.
The device contains 'zfs_member' signature and it may be removed by a write command. See sfdisk(8) man page and --wipe option for more details.

Created a new partition 1 of type 'Linux' and of size 931.5 GiB.
Partition #1 contains a zfs_member signature.

Do you want to remove the signature? [Y]es/[N]o: 

After stating "No" it outputs:

Do you want to remove the signature? [Y]es/[N]o: N
   /dev/sdk1 :         2048   1953525167 (931.5G) Linux

I did ask to not type anything else, but you chose the right option. Now type write and enter to write changes and post zpool import output once more

3 hours ago, JorgeB said:

I did ask to not type anything else, but you chose the right option. Now type write and enter to write changes and post zpool import output once more

Here we go:

root@Unraid:~# zpool import
  pool: cache
    id: 12246082903798894337
 state: UNAVAIL
status: One or more devices contains corrupted data.
action: The pool cannot be imported due to damaged devices or data.
   see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-5E
config:

        cache       UNAVAIL  insufficient replicas
          raidz1-0  UNAVAIL  insufficient replicas
            sdi1    ONLINE
            sdh1    ONLINE
            sdl1    UNAVAIL
            sdj1    FAULTED  corrupted data

it now seems looking for another drive... (now sdl instead of sdk)

But im not shure if it is worth in trying to bring back the old cache volume rather than simply creating a proper new one. Mover is scheduled to run once a day. cache is only configured to run cache->array. So all i have to loose is data since the last mover run right? If i am right it only should be the data of some docker container )bad enough but i can restore it as it should be plex DB and paperless DB, the data lays all directly on array)

The pool appears to be damaged beyond recovery; recommend reformatting it with the device you want to use.

alright. The i will edit the new pool to zfs and let it format the disks right? Thanks for the Help anyway. I will wait until this evening, as there is parity check running in its last 15% after the hard shutdown yesterday.

I will then report in a few days if the problem occurs again.

I also have a hunch where the error is coming from: A while back, I had to replace an SSD because I was getting an increasing number of CRC errors. I just replaced it through the main view drop downs and didn't change anything in the ZFS configurations (zpool cli). That must be what's causing the SSD to keep "disappearing."

7 minutes ago, alabiana said:

The i will edit the new pool to zfs and let it format the disks right?

Yep.

The time has come: the next SSD in the ZFS pool has been removed and is reporting “trimming.” The system has been running continuously since the pool was created.

  pool: ssd1pool
 state: DEGRADED
status: One or more devices have been removed.
	Sufficient replicas exist for the pool to continue functioning in a
	degraded state.
action: Online the device using zpool online' or replace the device with
	'zpool replace'.
  scan: scrub repaired 0B in 00:00:01 with 0 errors on Fri May 22 23:29:56 2026
config:

	NAME           STATE     READ WRITE CKSUM
	ssd1pool       DEGRADED     0     0     0
	  raidz1-0     DEGRADED     0     0     0
	    /dev/sdh1  ONLINE       0     0     0
	    /dev/sdi1  ONLINE       0     0     0
	    /dev/sdj1  ONLINE       0     0     0
	    /dev/sdk1  REMOVED      0     0     0  (trimming)

errors: No known data errors

in lsblk the ssd disappeared.

unraid-diagnostics-20260526-0656.zip

Syslog is being spammed with HBA issues:

May 26 05:07:21 Unraid kernel: mpt3sas_cm0 fault info from func: mpt3sas_base_make_ioc_ready

May 26 05:07:21 Unraid kernel: mpt3sas_cm0: fault_state(0x6004)!

May 26 05:07:21 Unraid kernel: mpt3sas_cm0: sending diag reset !!

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: diag reset: SUCCESS

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: CurrentHostPageSize is 0: Setting default host page size to 4k

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: basedisplay_fwpkg_version: complete

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: LSISAS3224: FWVersion(09.00.100.00), ChipRevision(0x01)

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: Protocol=(Initiator,Target), Capabilities=(TLR,EEDP,Snapshot Buffer,Diag Trace Buffer,Task Set Full,NCQ)

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: sending port enable !!

May 26 05:08:13 Unraid kernel: mpt3sas_cm0: port enable: SUCCESS

May 26 05:08:13 Unraid kernel: mpt3sas_cm0: search for end-devices: start

Make sure it's well seated and properly cooled; you can also try a different PCIe slot.

On 5/26/2026 at 8:50 AM, JorgeB said:

Syslog is being spammed with HBA issues:

May 26 05:07:21 Unraid kernel: mpt3sas_cm0 fault info from func: mpt3sas_base_make_ioc_ready

May 26 05:07:21 Unraid kernel: mpt3sas_cm0: fault_state(0x6004)!

May 26 05:07:21 Unraid kernel: mpt3sas_cm0: sending diag reset !!

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: diag reset: SUCCESS

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: CurrentHostPageSize is 0: Setting default host page size to 4k

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: basedisplay_fwpkg_version: complete

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: LSISAS3224: FWVersion(09.00.100.00), ChipRevision(0x01)

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: Protocol=(Initiator,Target), Capabilities=(TLR,EEDP,Snapshot Buffer,Diag Trace Buffer,Task Set Full,NCQ)

May 26 05:07:22 Unraid kernel: mpt3sas_cm0: sending port enable !!

May 26 05:08:13 Unraid kernel: mpt3sas_cm0: port enable: SUCCESS

May 26 05:08:13 Unraid kernel: mpt3sas_cm0: search for end-devices: start

Make sure it's well seated and properly cooled; you can also try a different PCIe slot.

I have placed a small fan on the heatblock of the hba. I will see if it resolves this. The hba was from an old server and my server is an ATX Case. So it is possible there was not enough airflow on the hba card.

Okay i now have a a fan mounted on the hba heatblock. But tonight a ssd disappeared again. I will look for a cli tool for my hba to read out data from there.

unraid-diagnostics-20260601-0844.zip

Still same issue, it may be worth trying a different PCI slot, if you haven't yet, or a different HBA