[BUG] Updating the firewall in a talos cluster prevents you from seeing cluster details

[fernando-villalba]

Description

Updating the firewall in a talos cluster prevents you from seeing cluster details

Reproducing the issue

1. Using the following code

provider "civo" {
  region = "LON1"
}

resource "civo_firewall" "example" {
    name = "example-firewall"
    create_default_rules = true
    network_id = civo_network.example.id

}

resource "civo_firewall" "example2" {
    name = "example-firewall2"
    create_default_rules = true
    network_id = civo_network.example.id

}

resource "civo_network" "example" {
  label = "example-network1234"

}

# Query instance disk image
data "civo_disk_image" "debian" {
   filter {
        key = "name"
        values = ["debian-10"]
   }
}

resource "civo_kubernetes_cluster" "example" {
    name = "example-cluster"
    # applications = "argocd,linkerd:Linkerd with Dashboard & Jaeger"
    network_id = civo_network.example.id
    firewall_id = civo_firewall.example.id # change this after creation to civo_firewall.example2.id

    cluster_type = "talos"
    pools {
        label = "shop"
        size = "g4s.kube.small"
        node_count = 3
    }
}

2. If you modify and apply changes for a firewall after creation, you get the following error:

3. The cluster can no longer be accessed via the UI, getting the following:

4. And if you try to get details in the CLI, you get the following, better error:

At this point, I am not sure if there is a fix for this broken cluster, at least not without contacting support

5. Moreover, After this happens, terraform thinks the cluster does not exist any longer and tries to recreate it:

6. Even though it still does:

Acceptance Criteria

• Ensure that terraform cannot break showing the cluster
• Ensure that the API does not let you break the cluster like this!

[fernando-villalba]

This is an internal issue outside of the scope of terraform, closing here, but kept as reference to reproduce issue.