Inconsistence when ibm_container_addons fails and run apply again

[feliperfmarques]

Hi there,

I percepted an inconsistence when ibm_container_addons fails. Acctually, I have faced many timeout errors with this resource and despite receiving error the addons are installed correctly on cluster, but believe that is about other issue. Anyway, I got timeout error, like this:

Error: Error waiting for Enabling Addon () : timeout while waiting for state to become 'normal, warning, critical, available' (timeout: 20m0s)

  on main.tf line 74, in resource "ibm_container_addons" "iks_1_us_south_addons":
  74: resource "ibm_container_addons" "iks_1_us_south_addons" {

After, when I try to run apply my terraform code again, I got this error:

Error: Request failed with status code: 409, ServerErrorResponse: {"incidentID":"6043f4e89559f69b-GRU","code":"E4d8b","description":"An add-on with the same name already exists. To update your existing add-on to the latest version, see 'http://ibm.biz/addon-update'.","type":"Provisioning"}

  on main.tf line 74, in resource "ibm_container_addons" "iks_1_us_south_addons":
  74: resource "ibm_container_addons" "iks_1_us_south_addons" {

For solve this, I need to remove ibm_container_addons from code, but the addons are not destroyed by terraform. I believe that before apply again or destroy the resource should check if this addon already exists and update terraform state and after run action.

Terraform Version

Terraform v0.13.5
provider registry.terraform.io/ibm-cloud/ibm v1.16.1

Affected Resource(s)

• ibm_container_addons

Terraform Configuration Files

resource "ibm_container_addons" "iks_1_us_south_addons" {
  depends_on        = [ibm_container_vpc_cluster.iks_1_us_south] 
  cluster           = ibm_container_vpc_cluster.iks_1_us_south.id
  resource_group_id = var.globals.resource_group

  addons {
    name    = "cluster-autoscaler"
    version = "1.0.1"
  }
  addons {
    name    = "alb-oauth-proxy"
    version = "1.0.0"
  }
  addons {
    name    = "istio"
    version = "1.7"
  }
}

Expected Behavior

When ibm_container_addons fails, resource should to be able to run apply code again, installing just addons still not installed and update terraform state, same for destroy action.

Actual Behavior

When ibm_container_addons fails and run apply code again I got error because the resource tries to install addons that have already been installed and If I remove ibm_container_addons from code the apply works fine, but the addons are not destroyed because they are not in the terraform state.

Steps to Reproduce

1. terraform apply and simulate fails in ibm_container_addons step
2. terraform apply again

[feliperfmarques]

Thanks so much @hkantare and @kavya498 for working on solving this issue 🥳