eks_nodegroup - wait for deletion of both node groups (#1994) (#1995)

[PR #1994/5248e3ff backport][stable-7] eks_nodegroup - wait for deletion of both node groups

This is a backport of PR #1994 as merged into main (5248e3f).
SUMMARY
Fixes eks_nodegroup integration test.
Ensure eks_nodegroup waits for deletion when wait=True including when the node group was already in deletING state.
ISSUE TYPE

Feature Pull Request

COMPONENT NAME
eks_nodegroup
ADDITIONAL INFORMATION

Reviewed-by: Mark Chappell

changelogs/fragments/eks_nodegroup-integration-wait-delete.yml

@@ -0,0 +1,4 @@
+trivial:
+- eks_nodegroup - update integration test to wait for both nodegroups to be deleted.
+minor_changes:
+- eks_nodegroup - ensure wait also waits for deletion to complete when ``wait==True`` (https://github.com/ansible-collections/community.aws/pull/1994).

plugins/modules/eks_nodegroup.py

@@ -607,18 +607,29 @@ def delete_nodegroups(client, module):
     clusterName = module.params["cluster_name"]
     existing = get_nodegroup(client, module, name, clusterName)
     wait = module.params.get("wait")
-    if not existing or existing["status"] == "DELETING":
-        module.exit_json(changed=False, msg="Nodegroup not exists or in DELETING status.")
-    if not module.check_mode:
-        try:
-            client.delete_nodegroup(clusterName=clusterName, nodegroupName=name)
-        except (botocore.exceptions.BotoCoreError, botocore.exceptions.ClientError) as e:
-            module.fail_json_aws(e, msg=f"Couldn't delete Nodegroup {name}.")
 
+    if not existing:
+        module.exit_json(changed=False, msg="Nodegroup '{name}' does not exist")
+
+    if existing["status"] == "DELETING":
         if wait:
             wait_until(client, module, "nodegroup_deleted", name, clusterName)
+            module.exit_json(changed=False, msg="Nodegroup '{name}' deletion complete")
+        module.exit_json(changed=False, msg="Nodegroup '{name}' already in DELETING state")
+
+    if module.check_mode:
+        module.exit_json(changed=True, msg="Nodegroup '{name}' deletion would be started (check mode)")
+
+    try:
+        client.delete_nodegroup(clusterName=clusterName, nodegroupName=name)
+    except (botocore.exceptions.BotoCoreError, botocore.exceptions.ClientError) as e:
+        module.fail_json_aws(e, msg=f"Couldn't delete Nodegroup '{name}'.")
+
+    if wait:
+        wait_until(client, module, "nodegroup_deleted", name, clusterName)
+        module.exit_json(changed=True, msg="Nodegroup '{name}' deletion complete")
 
-    module.exit_json(changed=True)
+    module.exit_json(changed=True, msg="Nodegroup '{name}' deletion started")
 
 
 def get_nodegroup(client, module, nodegroup_name, cluster_name):

tests/integration/targets/eks_nodegroup/tasks/cleanup.yml

@@ -74,10 +74,10 @@
     state: absent
     vpc_id: '{{ setup_vpc.vpc.id}}'
   ignore_errors: 'yes'
-    
+
 - name: remove setup VPC
   ec2_vpc_net:
     cidr_block: 10.0.0.0/16
     state: absent
     name: '{{ resource_prefix }}_aws_eks'
-  ignore_errors: 'yes'
+  ignore_errors: 'yes'

tests/integration/targets/eks_nodegroup/tasks/dependecies.yml

@@ -2,7 +2,7 @@
 # This space was a copy by aws_eks_cluster integration test
 - name: ensure IAM instance role exists
   iam_role:
-    name: ansible-test-eks_cluster_role
+    name: ansible-test-{{ tiny_prefix }}-eks_nodegroup-cluster
     assume_role_policy_document: '{{ lookup(''file'',''eks-trust-policy.json'') }}'
     state: present
     create_instance_profile: 'no'
@@ -44,7 +44,7 @@
   community.aws.ec2_vpc_route_table:
     vpc_id: '{{ setup_vpc.vpc.id }}'
     tags:
-      Name: EKS
+      Name: "EKS-ng-{{ tiny_prefix }}"
     subnets: '{{ setup_subnets.results | map(attribute=''subnet.id'') }}'
     routes:
       - dest: 0.0.0.0/0
@@ -77,9 +77,9 @@
       - eks_create.name == eks_cluster_name
 
 # Dependecies to eks nodegroup
-- name: create IAM instance role 
+- name: create IAM instance role
   iam_role:
-    name: 'ansible-test-eks_nodegroup'
+    name: 'ansible-test-{{ tiny_prefix }}-eks_nodegroup-ng'
     assume_role_policy_document: '{{ lookup(''file'',''eks-nodegroup-trust-policy.json'') }}'
     state: present
     create_instance_profile: no

tests/integration/targets/eks_nodegroup/tasks/full_test.yml

@@ -445,7 +445,6 @@
     state: absent
     cluster_name: '{{ eks_cluster_name }}'
   register: eks_nodegroup_result
-  check_mode: True
 
 - name: check that eks_nodegroup is not changed (idempotency)
   assert:
@@ -578,9 +577,21 @@
     cluster_name: '{{ eks_cluster_name }}'
     wait: True
   register: eks_nodegroup_result
-  check_mode: True
 
 - name: check that eks_nodegroup is not changed (idempotency)
   assert:
     that:
-      - eks_nodegroup_result is not changed
+      - eks_nodegroup_result is not changed
+
+- name: wait for deletion of name_a nodegroup (idempotency)
+  eks_nodegroup:
+    name: '{{ eks_nodegroup_name_a }}'
+    state: absent
+    cluster_name: '{{ eks_cluster_name }}'
+    wait: True
+  register: eks_nodegroup_result
+
+- name: check that eks_nodegroup is not changed (idempotency)
+  assert:
+    that:
+      - eks_nodegroup_result is not changed