"memory.hugepages" / "cpu.affinity" have invalid values

[andrewcfitz]

I am using the provider to make a full clone of a template. Everything works correctly the first run, and the vm works as expected.

However, the next time I run a preview I am met with the following errors:

error: proxmoxve:VM/virtualMachine:VirtualMachine resource 'dev' has a problem: expected hugepages to be one of ["1024" "2" "any"], got . Examine values at 'dev.memory.hugepages'.
error: proxmoxve:VM/virtualMachine:VirtualMachine resource 'dev' has a problem: invalid value for affinity (must contain numbers or number ranges separated by ','). Examine values at 'dev.cpu.affinity'.

I've tried ignoring memory.hugepages and cpu.affinity, but that doesn't help. I am not setting any properties in memory or cpu at all.

Is this a problem with the pulumi provider or the terraform provider?

[muhlba91]

generally speaking, this looks like an upstream issue to me where the retrieved values are not mapped properly (anymore). i raised one such issue upstream in the past because it broke the pulumi provider. could you please raise this issue upstream as well?

however, there was a property which raised a similar issue for me once. my workaround was to set the property in pulumi to its default value. did you try if this would solve your issue intermediary?

[Lewis-software]

I'm having the same bug. No matter what value I set, it gives me back the same error

[muhlba91]

i suggest it's best to raise this in the upstream provider (https://github.com/bpg/terraform-provider-proxmox) and link it to here for visibility.
to me it looks like the typing of these arguments/values do not reflect as in what is expected. did you try setting the same in a Terraform program?