TASK ERROR: command "blkid" "-o" "export" "/dev/sda1" failed - status code: 2 - no error message

[leofabri]

I've just installed PBS and I'm trying to create a Directory on my 10TB Seagate IronWolf drive.

Unfortunately, right when I'm issuing the create command, this happens:
TASK ERROR: command "blkid" "-o" "export" "/dev/sda1" failed - status code: 2 - no error message

Obviously, if I look into Administration > Storage / Disks > Directory, the drive is not present because the configuration failed.

I also tried to manually set up the disk after cleaning it from any existing partition:

proxmox-backup-manager disk initialize sda
proxmox-backup-manager disk fs create store1 --disk sda --filesystem ext4 --add-datastore true

However, the problem is still the same.

What seems to be happening is that after I delete every partition from the disk, issuing blkid returns the entry /dev/sda as expected; after the proxmox-backup-manager disk fs create ... though, the entry is gone and the disk is only visible through fdisk -l.
I feel like I don't have enough understanding of what's happening behind the hood.

ZFS, on the other hand, seems to be working.

 

[dcsapak]

can you post the output of

lsblk

and dmesg after you tried to initialize the disk with an fs ?

 

[leofabri]

Sure, here is it

$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sda 8:0 0 9.1T 0 disk
sdb 8:16 1 465.8G 0 disk
├─sdb1 8:17 1 1007K 0 part
├─sdb2 8:18 1 512M 0 part /boot/efi
└─sdb3 8:19 1 465.3G 0 part
├─pbs-swap 253:0 0 7G 0 lvm [SWAP]
└─pbs-root 253:1 0 442.3G 0 lvm /

$ dmesg

see the file attached

 

[leofabri]

Do you guys have any clue about what could be the cause of the problem? dmesg doesn't seem to report any meaningful info about what's going on during the creation of the fs.

 

[dcsapak]

ok looking at the dmesg output it looks rather weird (no text after the 'sda:' part)
what are the hardware components? especially sata controller/hba?

 

[leofabri]

ok looking at the dmesg output it looks rather weird (no text after the 'sda:' part)
what are the hardware components? especially sata controller/hba?

Yes, there are no details about the problem.
However, I'm using quite a constrained hardware. The motherboard is an Asrock N3700-ITX and the CPU (integrated) is an Intel Pentium Processor N3700 as you can see here.

The SATA controller is the integrated one.

Is there any way to make this work on this kind of hardware?

 

[leofabri]

Um, this is so strange...
Now it's working but I don't know why

That's what I did:

# Note: The 10TB disk is now sdb instead of sda (I swapped some ports the other day).

fdisk /dev/sdb
d                      # deleted everything again
n                      # created a new GPT
w                      # wrote to the disk
mkfs.ext4 /dev/sdb     # made an ext4 signature manually

Storage > Disks > Select (/dev/sdb that appears to be unused & unmounted) > Initialize Disk With GPT (shouldn't be necessary)
Storage > Add Directory > Disk: /dev/sdb | Type: ext4 | Name: storage1 > Create > OK

I don't know what's going on here. Unfortunately, I forgot to look into dmesg to see what happened.
Obviously, I'm going to test this thoroughly but do you think I can trust its reliability?

 

[dcsapak]

well i'd check dmesg but if it seems to work now... it's ok i guess? just make sure you have proper and working backups (thats true in any case though)

 

[leofabri]

well i'd check dmesg but if it seems to work now... it's ok i guess? just make sure you have proper and working backups (thats true in any case though)

Thank you Domink. So far so good! I tried to backup & restore more than 70GB of VMs and it's finally working. My poor backup machine is hitting 97% of cpu usage during a task but it seems stable.

I'm still learning how to use it but I'm already in love with this backup solution. And the client is so useful!
This definitely deserves so much more attention