nsx_edge_transport_node deploy failure (root object absent)

[martinrohrbach]

Describe the bug

Followup on #1186 since the second part of the issue hasn't been fixed yet.

We still have problems creating new edges with our desired configuration. Given the following resource:

resource "nsxt_edge_transport_node" "edge" {
  description  = var.edge.display_name
  display_name = var.edge.display_name

  standard_host_switch {
    host_switch_name = "Overlay"
    ip_assignment {
      static_ip_pool = nsxt_policy_ip_pool.ip_pool_edge.realized_id
    }
    transport_zone_endpoint {
      transport_zone = data.nsxt_policy_transport_zone.policy_overlay_transport_zone.id
    }
    host_switch_profile = [data.nsxt_policy_uplink_host_switch_profile.uplink_host_switch_profile.id]
    pnic {
      device_name = "fp-eth0"
      uplink_name = "uplink"
    }
  }

  standard_host_switch {
    host_switch_name = "VLAN"
    ip_assignment {
      assigned_by_dhcp = true
    }
    transport_zone_endpoint {
      transport_zone = data.nsxt_policy_transport_zone.policy_vlan_transport_zone.id
    }

    host_switch_profile = [data.nsxt_policy_uplink_host_switch_profile.uplink_host_switch_profile.id]
    pnic {
      device_name = "fp-eth1"
      uplink_name = "uplink"
    }
  }

  deployment_config {
    form_factor = var.edge.form_factor
    node_user_settings {
      cli_password  = var.edge.deploy_creds
      root_password = var.edge.deploy_creds
    }
    vm_deployment_config {
      management_network_id = data.vsphere_network.edge_mgmt_network.id
      data_network_ids = [
        data.vsphere_network.edge_geneve_network.id,
        data.vsphere_network.edge_uplink_network.id
      ]
      compute_id = data.vsphere_compute_cluster.edge_compute_cluster.id
      storage_id = data.vsphere_datastore.datastore_edges.id
      vc_id      = data.nsxt_compute_manager.compute_mgr.id
      host_id    = data.vsphere_host.edge_deployment_target_host.id
      management_port_subnet {
        ip_addresses  = [var.edge.mgmt_ip]
        prefix_length = 24
      }
      default_gateway_address = [var.edge.mgmt_default_gateway]
    }
  }

  node_settings {
    hostname             = var.edge.host_name
    allow_ssh_root_login = true
    enable_ssh           = false
    dns_servers          = var.edge.dns_servers
    ntp_servers          = var.edge.ntp_servers
    search_domains       = var.edge.search_domains
    syslog_server {
      name      = var.edge.syslog_server
      log_level = "INFO"
      protocol  = "TCP"
      server    = var.edge.syslog_server
      port      = 514
    }
  }

  depends_on = [
    nsxt_policy_ip_pool_static_subnet.static_subnet_edge,
  ]
}

The edge is deployed, but terraform throws an error at the end and the edge is not persisted in the state:

Error: Provider produced inconsistent result after apply
When applying changes to nsxt_edge_transport_node.edge,
provider "provider[\"registry.terraform.io/vmware/nsxt\"]" produced an
unexpected new value: Root object was present, but now absent.
This is a bug in the provider, which should be reported in the provider's
own issue tracker.

Reproduction steps

1. Deploy nsx_edge_transport_node with the above config
2. Observe error(s)

Expected behavior

No errors on deploy.

Additional context

We're testing against NSX-T 4.1.2.3 with the newest provider 3.6.1.

[annakhm]

Hi @martinrohrbach, does this happen with clean state, or is this a result of apply on state that resulted from previous error (after upgrading the provider)? Thank you

[martinrohrbach]

@annakhm The repo had a clean apply before without any nsxt_edge_transport_node in the state. I then updated from provider 3.6.0 to to provider 3.6.1, added the nsxt_edge_transport_node and applied it which resulted in the error above. Does that answer the question? Or did I get the question wrong?

If you think this has to do with the previous state, I could test this again with a completely empty state, I'll see if I find some time for this on monday.

[martinrohrbach]

@annakhm Update: I just tested this with a blank terraform state, creation of the edge resource only.

Same result, edge is being created just fine, terraform (1.8.2) errors out with the error above, state is empty afterwards.

[annakhm]

Thank you, this helps!

[martinrohrbach]

@annakhm
I don't think this fix has made to a released version yet, any chance to get an update on when the next release is due?

[annakhm]

Hi @martinrohrbach, we expect next release to be before end of July