aws_eks_cluster: Add wait functionality (#42259)

* aws_eks_cluster: Improve output documentation

This data is already returned by the module, it just wasn't documented. These
fields are required for accessing the created Kubernetes API with e.g. the
k8s_raw module.

* aws_eks_cluster: Add wait functionality

This enables further cluster configuration once it's created and active.

20 minutes was chosen as an arbitrary default, so that if it takes longer than
the documented "usually less than 10 minutes" it's still likely to succeed.

* Correct security group name in aws_eks tests

* Improve teardown of aws_eks tests

Fix minor teardown issues. The `pause` step is a placeholder until
a waiter for `state: absent`

lib/ansible/module_utils/aws/waiters.py

@@ -179,6 +179,30 @@ waf_data = {
     }
 }
 
+eks_data = {
+    "version": 2,
+    "waiters": {
+        "ClusterActive": {
+            "delay": 20,
+            "maxAttempts": 60,
+            "operation": "DescribeCluster",
+            "acceptors": [
+                {
+                    "state": "success",
+                    "matcher": "path",
+                    "argument": "cluster.status",
+                    "expected": "ACTIVE"
+                },
+                {
+                    "state": "retry",
+                    "matcher": "error",
+                    "expected": "ResourceNotFoundException"
+                }
+            ]
+        }
+    }
+}
+
 
 def ec2_model(name):
     ec2_models = core_waiter.WaiterModel(waiter_config=ec2_data)
@@ -190,6 +214,11 @@ def waf_model(name):
     return waf_models.get_waiter(name)
 
 
+def eks_model(name):
+    eks_models = core_waiter.WaiterModel(waiter_config=eks_data)
+    return eks_models.get_waiter(name)
+
+
 waiters_by_name = {
     ('EC2', 'route_table_exists'): lambda ec2: core_waiter.Waiter(
         'route_table_exists',
@@ -251,6 +280,12 @@ waiters_by_name = {
         core_waiter.NormalizedOperationMethod(
             waf.get_change_token_status
         )),
+    ('EKS', 'cluster_active'): lambda eks: core_waiter.Waiter(
+        'cluster_active',
+        eks_model('ClusterActive'),
+        core_waiter.NormalizedOperationMethod(
+            eks.describe_cluster
+        )),
 }