LVM storage pool on IncusOS

IncusOS
1

I am trying to create a cached LVM storage pool on IncusOS. According to this discussion I can do that by reusing a manually created LVM storage pool by setting lvm.vg.force_reuse=true.

First, I am wondering if the LVM service should be enabled on IncusOS to create LVM storage pools? According to the documentation:

The LVM service allows configuring of a clustered LVM storage backend.

I am not using any cluster. But upon creation of the pool I get:

~ % incus storage create backup lvm source=vg_test lvm.vg.force_reuse=true lvm.use_thinpool=true

Error: IncusOS service "lvm" isn't currently enabled

After enabling the LVM service the storage pool gets created successfully and seems functional even if I disable this service later.

Next, upon reboot, I notice the following error on IncusOS log:

[2026/08/13 14:55:53 CEST] lvm: modprobe: ERROR: ctx=0x55f5483ac2a0 path=/lib/modules/7.1.8-zabbly+/kernel/drivers/md/persistent-data/dm-persistent-data.ko error=No such file or directory
[2026/08/13 14:55:53 CEST] lvm: modprobe: ERROR: ctx=0x55f5483ac2a0 path=/lib/modules/7.1.8-zabbly+/kernel/drivers/md/persistent-data/dm-persistent-data.ko error=No such file or directory
[2026/08/13 14:55:53 CEST] lvm: modprobe: ERROR: could not insert 'dm_thin_pool': Unknown symbol in module, or unknown parameter (see dmesg)
[2026/08/13 14:55:53 CEST] lvm:   /sbin/modprobe failed: 1
[2026/08/13 14:55:53 CEST] lvm:   Can't process LV vg_test/test_pool: thin-pool target support missing from kernel?
[2026/08/13 14:55:53 CEST] lvm:   0 logical volume(s) in volume group "vg_test" now active
[2026/08/13 14:55:53 CEST] systemd: lvm-activate-vg_test.service: Main process exited, code=exited, status=5/NOTINSTALLED
[2026/08/13 14:55:53 CEST] systemd: lvm-activate-vg_test.service: Failed with result 'exit-code'.

I created the volume group and thin volume in a separate VM lvm-worker by passing the disk path from the host:

root@lvm-worker:~# vgcreate vg_test /dev/sdb
  Volume group "vg_test" successfully created
root@lvm-worker:~# lvcreate --thinpool test_pool vg_test -L 10G
  Thin pool volume with chunk size 64.00 KiB can address at most <15.88 TiB of data.
  Logical volume "test_pool" created.

The storage pool seems to work fine, but I am wondering if the modprobe errors point to some missing module dependencies to dm-persistent-data.ko in the Zabbly kernel?

I had similar errors when trying to use a volume with cache devices although the storage pool worked again fine in Incus:

[2026/08/13 13:41:17 CEST] lvm: modprobe: ERROR: ctx=0x559cf16682a0 path=/lib/modules/7.1.8-zabbly+/kernel/drivers/md/persistent-data/dm-persistent-data.ko error=No such file or directory
[2026/08/13 13:41:17 CEST] lvm: modprobe: ERROR: ctx=0x559cf16682a0 path=/lib/modules/7.1.8-zabbly+/kernel/drivers/md/persistent-data/dm-persistent-data.ko error=No such file or directory
[2026/08/13 13:41:17 CEST] lvm: modprobe: ERROR: could not insert 'dm_cache': Unknown symbol in module, or unknown parameter (see dmesg)
[2026/08/13 13:41:17 CEST] lvm:   /sbin/modprobe failed: 1
[2026/08/13 13:41:17 CEST] lvm:   Can't process LV vg_backup/tm_backup: cache target support missing from kernel?
[2026/08/13 13:41:17 CEST] lvm:   0 logical volume(s) in volume group "vg_backup" now active
[2026/08/13 13:41:17 CEST] audit: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=unconfined msg='unit=lvm-activate-vg_backup comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=failed'
[2026/08/13 13:41:17 CEST] systemd: lvm-activate-vg_backup.service: Main process exited, code=exited, status=5/NOTINSTALLED
[2026/08/13 13:41:17 CEST] systemd: lvm-activate-vg_backup.service: Failed with result 'exit-code'.

My system:

IncusOS Version: 202608102114
Installed applications
incus 202608102114

2

You probably want to check the full boot log. You may be looking at very early boot (initrd) entries which run with a much narrower set of modules than the fully booted system.

3

Indeed, the messages are quite early in the boot. Right before:

[2026/08/13 16:03:51 CEST] systemd: Finished initrd-multipath.service - Configure basic multipath.

Later in the boot the device is reported to be online.

[2026/08/13 16:03:51 CEST] lvm: PV /dev/sda online, VG vg_backup is complete.

So I guess the messages are benign. It just caught my eye because of ERROR in all caps.

What about the LVM service? Should it be enabled?

4

The LVM service is really mostly there for LVM cluster, though I believe we also use it to keep /etc/lvm/ clean which is why Incus requires it for pool creation.