Din Mušić
Din Mušić
This seems like a bug. I was able to reproduce the issue, so I'll investigate what is going on.
Yes, thanks
Hi could you share the configuration of the entire `lxd_profile`/`lxd_instance`?
It does not work in the profile because `initial.*` options are disk device options. Here is an example where `initial.zfs.block_mode` is set to the root disk device within a profile...
> we do not wish to manage lxd_storage_pools within terraform The `initial.*` settings may prove useful in such case, as they allow configuring new instances with different volume settings without...
I'm not sure I understand the question. Attaching custom volumes should work both with pre-existing storage pools, and storage pools that created via TF
Ahh, got you. You can use an existing storage pool (not managed by Terraform) by referencing it by its name: ```sh # Create storage pool. lxc storage create tmp zfs...
Hmm, I've tried both with VMs and containers with the following example: ``` lxc storage create local zfs ``` ```hcl resource "lxd_volume" "dockerzvol" { name = "dockerzvol" pool = "local"...
> When running with initial.zfs.block_mode = true in the device properties Initial settings only apply to instance root disk volume, not the custom volume. > And when specifying content_type =...
> Could not locate a zvol for rpool/lxd/containers/int-dev-manager01 Yes, this was an issue in LXD where it could happen that udev rules were not applied in time (especially on loaded...