How to use '-force' option with 'zpool create' on IncusOS?

IncusOS
, ,
1

Hey there, i want to create a raidz1 pool with 3x1TB nvme drives.
They differ a bit in size so the zpool creation is failing.

the config i added:

config:
  pools:
  - name: “nvmepool”
    type: “zfs-raidz1”
    devices:
    - “/dev/disk/by-id/nvme-KINGSTON_OM8PGP41024Q-A0_50026B7382F42193”
    - “/dev/disk/by-id/nvme-SAMSUNG_MZVLB1T0HALR-000L7_S3TPNE0JC09446”
    - “/dev/disk/by-id/nvme-WDS100T3X0C-00SJG0_212103A00297”

❯ incus admin os system storage edit
WARNING: The IncusOS API and configuration is subject to change

Config parsing error: Failed to run: zpool create -o ashift=12 -O mountpoint=none -O encryption=aes-256-gcm -O keyformat=raw -O keylocation=file:///var/lib/incus-os/zpool.nvmepool.key nvmepool raidz1 /dev/disk/by-id/nvme-KINGSTON_OM8PGP41024Q-A0_50026B7382F42193 /dev/disk/by-i
d/nvme-SAMSUNG_MZVLB1T0HALR-000L7_S3TPNE0JC09446 /dev/disk/by-id/nvme-WDS100T3X0C-00SJG0_212103A00297: exit status 1 (invalid vdev specification
use '-f' to override the following errors:
raidz contains devices of different sizes)
Press enter to open the editor again or ctrl+c to abort change

Thanks for any help upfront

2

incus admin os system storage wipe-drive -d '{"id": "/dev/disk/by-id/nvme-KINGSTON_OM8PGP41024Q-A0_50026B7382F42193"}' can be used to wipe clean a particular drive. Repeat for all of the drives that may contain data, then try creating the pool again.

3

Thanks for your hint but i wiped the disks already before. I assume that this error is the result of different drive sizes due to nvme vendor mix.

Device: /dev/nvme0, KINGSTON OM8PGP41024Q-A0, S/N:XXXXXXXX, FW:ELFK0S.6, 1.02 TB
Device: /dev/nvme1, WDS100T3X0C-00SJG0, S/N:XXXXXXXX, FW:111130WD, 1.00 TB
Device: /dev/nvme2, SAMSUNG MZVLB1T0HALR-000L7, S/N:XXXXXXXX, FW:5L2QEXA7, 1.02 TB

4

Ah yeah, that will be slightly harder to support as we don’t want to offer a universal force flag as zpool create -f could do some very bad things if passed the wrong devices by accident… It would be nice if there was more granularity to the flag so we could allow the mis-matching sizes without allowing the force creation on in-use disks…

@gibmat maybe the solution here is to have a flag allowing the size mismatch, then attempt a normal zpool create, check that it’s failing because of the mismatch (and only because of that), then retry with --force?