Good afternoon.
I have a problem migrating from lxd to incus. I installed incus and ran lxd-to-incus, but it failed at “starting the target server”. Now, when I try to start incus, incusd fails with the following error messages:
incus[619]: time="2023-12-29T12:58:26+01:00" level=error msg="Failed mounting storage pool" err="Failed to run: zpool import -f -d /var/lib/incus/disks default: exit status 1 (cannot import 'default': no such pool available)" pool=default
incus[619]: time="2023-12-29T12:58:26+01:00" level=error msg="Failed to start the daemon" err="Failed applying patch \"storage_delete_old_snapshot_records\": Unvailable storage pools: [default]"
The lxd pool gets imported correctly, though, but lxd fails to start.
I guess I need to delete the default pool, but how can I do that?
Edit : I managed to get incus to start by creating a “default” pool on a vdev. I was then able to remove the default pool using incus storage.
Hmm, okay, so that means that there was a default
storage pool on the LXD side which was apparently configured to use a loop file in /var/snap/lxd/common/lxd/disks
, but following the move to /var/lib/incus/disks
, the loop couldn’t be found anymore.
That’s a bit odd.
You mention you managed to unblock this by temporarily creating the zpool, but does that mean that the default
pool was somehow unused/half-deleted?
What’s the state of your system now? Is everything working properly on Incus?
Right on point (although I don’t have any idea where that default storage pool might have come from). It seems there was some sort of a ghost config lying around without my being aware of it.
There was a .img file lying around in /var/lib/incus/disks which incus reported as being the source of default by incus storage show, although as I was unaware of that, I created my dummy zpool elsewhere.
Anyhow, everything is working properly now. Many thanks for the support.