[SOLVED] ZFS Issue : "protection error" thrown by the kernel

[Alexandre91]

Hi everyone !

I have a ProLiant ML150 Gen9 with ProxMox installed on it. I have 4 Seagate disks ST80000NM0075 mounted in raidz1. When I launch a "zpool status", I have these issues :

root@pve:~# zpool status
  pool: donnees
 state: DEGRADED
status: One or more devices are faulted in response to persistent errors.
    Sufficient replicas exist for the pool to continue functioning in a
    degraded state.
action: Replace the faulted device, or use 'zpool clear' to mark the device
    repaired.
  scan: resilvered 4.60G in 00:01:50 with 0 errors on Sun Oct 15 18:42:48 2023
config:

    NAME                        STATE     READ WRITE CKSUM
    donnees                     DEGRADED     0     0     0
      raidz1-0                  DEGRADED     7     0     0
        scsi-35000c500af376a63  ONLINE       0     0     0
        scsi-35000c500af20f9cb  FAULTED     13     0     0  too many errors
        scsi-35000c500af378d2b  ONLINE       0     0     0
        scsi-35000c500af379f63  DEGRADED   340     0     0  too many errors

errors: No known data errors

Furthermore, when I inspect errors with dmesg, I have ones :

root@pve:~# dmesg
[ 3388.618067] zio pool=donnees vdev=/dev/disk/by-id/scsi-35000c500af379f63-part1 error=84 type=1 offset=3377562501120 size=942080 flags=40080ca8
[ 3409.754155] sd 0:0:4:0: [sdd] tag#33 FAILED Result: hostbyte=DID_SOFT_ERROR driverbyte=DRIVER_OK cmd_age=0s
[ 3409.754164] sd 0:0:4:0: [sdd] tag#33 Sense Key : Aborted Command [current] [descriptor]
[ 3409.754168] sd 0:0:4:0: [sdd] tag#33 Add. Sense: Logical block guard check failed
[ 3409.754172] sd 0:0:4:0: [sdd] tag#33 CDB: Read(16) 88 00 00 00 00 01 9b 2d 84 d0 00 00 00 a8 00 00
[ 3409.754174] protection error, dev sdd, sector 6898418960 op 0x0:(READ) flags 0x700 phys_seg 3 prio class 2
[ 3409.754557] zio pool=donnees vdev=/dev/disk/by-id/scsi-35000c500af379f63-part1 error=84 type=1 offset=3531989426176 size=86016 flags=40080ca8
[ 3426.572376] sd 0:0:4:0: [sdd] tag#39 FAILED Result: hostbyte=DID_SOFT_ERROR driverbyte=DRIVER_OK cmd_age=0s
[ 3426.572387] sd 0:0:4:0: [sdd] tag#39 Sense Key : Aborted Command [current] [descriptor]
[ 3426.572391] sd 0:0:4:0: [sdd] tag#39 Add. Sense: Logical block guard check failed
[ 3426.572396] sd 0:0:4:0: [sdd] tag#39 CDB: Read(16) 88 00 00 00 00 01 a5 2f 1f 60 00 00 00 a8 00 00
[ 3426.572399] protection error, dev sdd, sector 7066296200 op 0x0:(READ) flags 0x700 phys_seg 3 prio class 2
[ 3426.572856] zio pool=donnees vdev=/dev/disk/by-id/scsi-35000c500af379f63-part1 error=84 type=1 offset=3617942585344 size=86016 flags=40080ca8
[ 3427.039858] sd 0:0:4:0: [sdd] tag#25 FAILED Result: hostbyte=DID_SOFT_ERROR driverbyte=DRIVER_OK cmd_age=0s
[ 3427.039868] sd 0:0:4:0: [sdd] tag#25 Sense Key : Aborted Command [current] [descriptor]
[ 3427.039872] sd 0:0:4:0: [sdd] tag#25 Add. Sense: Logical block guard check failed
[ 3427.039877] sd 0:0:4:0: [sdd] tag#25 CDB: Read(16) 88 00 00 00 00 01 a5 30 ac f8 00 00 02 b8 00 00
[ 3427.039880] protection error, dev sdd, sector 7066398448 op 0x0:(READ) flags 0x700 phys_seg 5 prio class 2
[ 3427.040332] zio pool=donnees vdev=/dev/disk/by-id/scsi-35000c500af379f63-part1 error=84 type=1 offset=3617994698752 size=356352 flags=40080ca8
[ 3431.357979] sd 0:0:4:0: [sdd] tag#36 FAILED Result: hostbyte=DID_SOFT_ERROR driverbyte=DRIVER_OK cmd_age=0s
[ 3431.357988] sd 0:0:4:0: [sdd] tag#36 Sense Key : Aborted Command [current] [descriptor]
[ 3431.357991] sd 0:0:4:0: [sdd] tag#36 Add. Sense: Logical block guard check failed
[ 3431.357995] sd 0:0:4:0: [sdd] tag#36 CDB: Read(16) 88 00 00 00 00 01 ab 2d f7 40 00 00 00 a8 00 00
[ 3431.357997] protection error, dev sdd, sector 7166883712 op 0x0:(READ) flags 0x700 phys_seg 3 prio class 2
[ 3431.358371] zio pool=donnees vdev=/dev/disk/by-id/scsi-35000c500af379f63-part1 error=84 type=1 offset=3669443379200 size=86016 flags=40080ca8
[ 3431.658042] sd 0:0:4:0: [sdd] tag#20 FAILED Result: hostbyte=DID_SOFT_ERROR driverbyte=DRIVER_OK cmd_age=0s
[ 3431.658052] sd 0:0:4:0: [sdd] tag#20 Sense Key : Aborted Command [current] [descriptor]
[ 3431.658055] sd 0:0:4:0: [sdd] tag#20 Add. Sense: Logical block guard check failed
[ 3431.658059] sd 0:0:4:0: [sdd] tag#20 CDB: Read(16) 88 00 00 00 00 01 ab 2e 5f 28 00 00 07 e8 00 00
[ 3431.658061] protection error, dev sdd, sector 7166911790 op 0x0:(READ) flags 0x700 phys_seg 11 prio class 2
[ 3431.658440] zio pool=donnees vdev=/dev/disk/by-id/scsi-35000c500af379f63-part1 error=84 type=1 offset=3669456998400 size=1036288 flags=40080ca8
[ 3431.730227] sd 0:0:2:0: [sdb] tag#42 FAILED Result: hostbyte=DID_SOFT_ERROR driverbyte=DRIVER_OK cmd_age=0s
[ 3431.730235] sd 0:0:2:0: [sdb] tag#42 Sense Key : Aborted Command [current] [descriptor]
[ 3431.730239] sd 0:0:2:0: [sdb] tag#42 Add. Sense: Logical block guard check failed
[ 3431.730242] sd 0:0:2:0: [sdb] tag#42 CDB: Read(16) 88 00 00 00 00 01 ab 2e 5f 30 00 00 04 18 00 00
[ 3431.730244] protection error, dev sdb, sector 7166910734 op 0x0:(READ) flags 0x700 phys_seg 13 prio class 2
[ 3431.730616] zio pool=donnees vdev=/dev/disk/by-id/scsi-35000c500af20f9cb-part1 error=84 type=1 offset=3669457002496 size=536576 flags=40080ca8

The problem is there is no SMART errors on my SAS disks. The HBA card is a H240 with the latest firmware (7.20).

Could you help me to determine where is the problem, and what is the meaning of "protection error", because I haven't found anything on the web ?

Best regards.

 

[leesteken]

Could you help me to determine where is the problem, and what is the meaning of "protection error", because I haven't found anything on the web ?

My web search suggested that it are remnants of previous RAID controller configuration on the drives and that a Seagate SeaTools Basic Format will fix it (and probably wipe the drives) : https://forums.servethehome.com/index.php?threads/what-is-wrong-with-my-drives.32376/#post-338658

EDIT: I don't think this is a ZFS issue at all as the drives or the HBA are actually reporting the errors, and I'm convinced you would get then when reading from the raw drives without ZFS.

 

[Alexandre91]

My web search suggested that it are remnants of previous RAID controller configuration on the drives and that a Seagate SeaTools Basic Format will fix it (and probably wipe the drives) : https://forums.servethehome.com/index.php?threads/what-is-wrong-with-my-drives.32376/#post-338658

Yep, on another server, these disks were used with a Perc H700 RAID controller. I will try to do a basic format with SeaTools and give news after that !

EDIT: I don't think this is a ZFS issue at all as the drives or the HBA are actually reporting the errors, and I'm convinced you would get then when reading from the raw drives without ZFS.

No, when I try to write data on these without ZFS, I have same issues...

 

[Alexandre91]

Hi!

I used the tool "SeaChest_Erase" from SeaTools, and it seems it solved my problem!
If someone have the same problem, he can tell me without problem.

Thank you very much,
Alexandre

 

[leesteken]

I used the tool "SeaChest_Erase" from SeaTools, and it seems it solved my problem!

Please mark this thread as Solved by editing your first post, so other people will find this solution quicker.

 

[Alexandre91]

Please mark this thread as Solved by editing your first post, so other people will find this solution quicker.

OK, I added the tag "SOLVED" before the topic name...