Allow Updating machine_type and Disk Size Without Resource Recreation

[PeKaHaMup]

Problem description

Currently, modifying the machine_type or the disk size of a VM using the STACKIT Terraform Provider results in the resource being recreated.

This behavior causes avoidable downtime and operational overhead, especially in production environments where vertical scaling (CPU/RAM increase) or disk expansion is required without changing the overall VM identity.

Proposed solution

Enable in-place updates of the following attributes, where supported by the underlying STACKIT API:

machine_type (CPU/RAM resizing)

Disk size (volume expansion)

The provider should perform an update operation instead of forcing resource recreation when these attributes change and the platform supports resizing.

resource "stackit_compute_instance" "example" {
name = "example-vm"
machine_type = "c2-standard-4" # should be updatable without recreation

boot_volume {
size = 100 # should be expandable without recreation
}

network {
network_id = stackit_network.example.id
}
}

Alternative solutions (optional)

Additional information

Supporting in-place updates for machine_type and disk size would significantly improve operational flexibility and reduce service interruptions during routine scaling operations.

[alexrothgit]

We stumbled over the same issue and figured that explicitly setting one of the boot disk parameter does fix this issue.

Setting delete_on_termination to true or false, does change the behavior.
Which is kind of weird. Whats the value if the parameter is unset?

Could you confirm this?

@Stackit: is this intended?

Thanks.