[SOLVED] Erroneous sense key warnings on SAS SSD

[Mastsrl]

Hello,
i've just installed PVE 8.3 on a Dell R730 with several SAS SSDs and i've noticed the following erroneous warnings on kernel log for every drive:

[ 2038.529350] sd 0:0:6:0: [sdd] tag#104 Sense Key : Recovered Error [current]
[ 2038.529400] sd 0:0:6:0: [sdd] tag#104 Add. Sense: Grown defect list not found

That repeats everytime proxmox reads the smart status via smartctl, which also makes the "Non-medium error count" of the drives increase by 1.

Googling for this i've seen these errors being reported as far back as 5+ years ago, there are some workarounds that require manual patching of smartctl files and re-compiling:
https://www.smartmontools.org/ticket/1448

https://pv-tech.eu/posts/smartctl-causing-errors-in-dmesg-with-sas-ssd-drives/

Since it seems the issue has been abandoned by the smartmontools team, any chance of being done by proxmox?, recompiling the smartmontool in vpe is a little bit outside my expertise

Also i've noticed that the smartctl version included in PVE is ancient (7.3) when 7.4 has been out since AUGUST of 2023(but the changelog of 7.4 makes no mention of this bug so it would not make a difference for this)...

 

[elmarconi]

Watching this as this just popped up on my fresh installed hPE ProLiant DL360 Gen10 Plus/ProLiant DL360 Gen10 Plus, BIOS U46 10/04/2024

[ 9852.955245] sd 0:2:6:0: [sdf] tag#1901 Sense Key : Recovered Error [current] 
[ 9852.955257] sd 0:2:6:0: [sdf] tag#1901 Add. Sense: Defect list not found
[ 9853.106658] sd 0:2:5:0: [sde] tag#7971 Sense Key : Recovered Error [current] 
[ 9853.106669] sd 0:2:5:0: [sde] tag#7971 Add. Sense: Defect list not found

root@pve4:~# fdisk -l /dev/sdf
Disk /dev/sdf: 1.75 TiB, 1920383410176 bytes, 3750748848 sectors
Disk model: VO001920RWUFE

 

[Mastsrl]

Watching this as this just popped up on my fresh installed hPE ProLiant DL360 Gen10 Plus/ProLiant DL360 Gen10 Plus, BIOS U46 10/04/2024

[ 9852.955245] sd 0:2:6:0: [sdf] tag#1901 Sense Key : Recovered Error [current]
[ 9852.955257] sd 0:2:6:0: [sdf] tag#1901 Add. Sense: Defect list not found
[ 9853.106658] sd 0:2:5:0: [sde] tag#7971 Sense Key : Recovered Error [current]
[ 9853.106669] sd 0:2:5:0: [sde] tag#7971 Add. Sense: Defect list not found

root@pve4:~# fdisk -l /dev/sdf
Disk /dev/sdf: 1.75 TiB, 1920383410176 bytes, 3750748848 sectors
Disk model: VO001920RWUFE

I manually updated smartmontools to latest 7.4 stable and the erroneous errors seem to have vanished from the kernel log.

To do this you have to edit sources.list:

editor /etc/apt/sources.list

and add the bookworm-backports repo to the line with "main" so it ends up like(it will only affect packages installed manually from the backports, it won't update any other already-installed package with "unstable" updates):

deb http://deb.debian.org/debian bookworm-backports main

and then:

apt install -t bookworm-backports smartmontools

that should get you to smarmontools 7.4