I have one RaidZ1 pool with 3 HDDs, recently one is supposedly dead (usually after rebooting it kinda works for some time), anyway, with this pool suspended Unraid doesn't turn off or reboot, it stops at "Unmounting local file systems:" requiring manually to shut it down.

 

also this is the zpool status:

 

  pool: nasa
 state: SUSPENDED
status: One or more devices are faulted in response to IO failures.
action: Make sure the affected devices are connected, then run 'zpool clear'.
   see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-HC
  scan: resilvered 576M in 00:00:07 with 0 errors on Sun Jan  5 19:14:50 2025
config:

        NAME        STATE     READ WRITE CKSUM
        nasa        DEGRADED     0     0     0
          raidz1-0  DEGRADED     7   150     0
            sde1    REMOVED      0     0     0
            sdd1    ONLINE      12   182     0
            sdb1    ONLINE       8     0     0

errors: 70 data errors, use '-v' for a list

 

1 hour ago, Chriexpe said:

pool suspended

After this happens, you will have to force a reboot, there were errors with 2 disks, I assume at the same time for the pool to get suspended, it's a zfs issue, not way around it AFAIK, reboot and run a scrub, but if there are two bad disks, the pool will suspend again.

5 hours ago, JorgeB said:

you will have to force a reboot

 

I'm aware, but shouldn't Unraid be more "resilient" to it (like not depending on disks)? It hangs during reboot/shutdown and I have to manually reset the server (thank god I have PiKVM).

Also, please take a look at diagnostics, specifically at syslog-previous as Unraid just mysteriously died some hours after the ZFS pool getting suspended, there wasn't any log or video output at all, I've read even some error with my SAS card too.

 

 

Interestingly enough, after a fresh reboot everything works fine (note that there is a NVR recording to this pool 24/7), tho one disk (sdd) lags behind in R/W.

 

 

Not only that but looking at the logs, that disk (sdd) has some more log that I couldn't comprehend right.

 

Do you have any idea on what could be? If it's actually one disk failing I may be able to get warranty on it (all HDDs are identical, Seagate Exos 7e8 8TB).

Some years ago I also got some I/O error and changing the SAS Sata cable solved it (I'm even considering getting one from Aliexpress), but even after swapping to the old cable I also got this issue on (possibly) the same HDD.

tower-diagnostics-20250106-2111.zip

Edited January 7, 20251 yr by Chriexpe

7 hours ago, Chriexpe said:

I'm aware, but shouldn't Unraid be more "resilient" to it (like not depending on disks)?

Like mentioned it's a zfs issue, it will happen with any OS if a pool gets suspended, requires a forced shutdown/reboot  

 

I'll take a look at the diags later, on a phone now.

 

 

There are new CRC errors for the 3rd pool disk, replace the SATA cable and run a scrub.