rds_cluster: add support for removing cluster from global db

changelogs/fragments/1705-rds_cluster-add-support-remove-cluster-from-global-db.yml

@@ -0,0 +1,3 @@
+---
+minor_changes:
+- rds_cluster - Add support for removing cluster from global db (https://github.com/ansible-collections/amazon.aws/pull/1705).

plugins/module_utils/rds.py

@@ -123,6 +123,8 @@ def get_rds_method_attribute(method_name, module):
             waiter = "role_disassociated"
         elif method_name == "promote_read_replica":
             waiter = "read_replica_promoted"
+        elif method_name == "db_cluster_promoting":
+            waiter = "db_cluster_promoting"
         else:
             waiter = "db_instance_available"
         # Handle retry codes

plugins/module_utils/waiters.py

@@ -547,6 +547,19 @@
 rds_data = {
     "version": 2,
     "waiters": {
+        "DBClusterPromoting": {
+            "delay": 5,
+            "maxAttempts": 60,
+            "operation": "DescribeDBClusters",
+            "acceptors": [
+                {
+                    "state": "success",
+                    "matcher": "pathAll",
+                    "argument": "DBClusters[].Status",
+                    "expected": "promoting",
+                },
+            ],
+        },
         "DBInstanceStopped": {
             "delay": 20,
             "maxAttempts": 60,
@@ -911,6 +924,11 @@ def route53_model(name):
         elbv2_model("LoadBalancersDeleted"),
         core_waiter.NormalizedOperationMethod(elbv2.describe_load_balancers),
     ),
+    ("RDS", "db_cluster_promoting"): lambda rds: core_waiter.Waiter(
+        "db_cluster_promoting",
+        rds_model("DBClusterPromoting"),
+        core_waiter.NormalizedOperationMethod(rds.describe_db_clusters),
+    ),
     ("RDS", "db_instance_stopped"): lambda rds: core_waiter.Waiter(
         "db_instance_stopped",
         rds_model("DBInstanceStopped"),

plugins/modules/rds_cluster.py

@@ -305,6 +305,13 @@
         aliases:
           - maintenance_window
         type: str
+    remove_from_global_db:
+        description:
+          - If set to C(true), the cluster will be removed from global DB.
+          - Parameters I(global_cluster_identifier), I(db_cluster_identifier) must be specified when I(remove_from_global_db=true).
+        type: bool
+        required: False
+        version_added: 6.5.0
     replication_source_identifier:
         description:
           - The Amazon Resource Name (ARN) of the source DB instance or DB cluster if this DB cluster is created as a Read Replica.
@@ -463,6 +470,42 @@
     engine: aurora-postgresql
     state: present
     db_instance_class: 'db.t3.medium'
+
+- name: Remove a cluster from global DB (do not delete)
+  amazon.aws.rds_cluster:
+    db_cluster_identifier: '{{ cluster_id }}'
+    global_cluster_identifier: '{{ global_cluster_id }}'
+    remove_from_global_db: true
+
+- name: Remove a cluster from global DB and Delete without creating a final snapshot
+  amazon.aws.rds_cluster:
+    engine: aurora
+    password: "{{ password }}"
+    username: "{{ username }}"
+    cluster_id: "{{ cluster_id }}"
+    skip_final_snapshot: true
+    remove_from_global_db: true
+    wait: true
+    state: absent
+
+- name: Update cluster port and WAIT for remove secondary DB cluster from global DB to complete
+  amazon.aws.rds_cluster:
+    db_cluster_identifier: "{{ secondary_cluster_name }}"
+    global_cluster_identifier: "{{ global_cluster_name }}"
+    remove_from_global_db: true
+    state: present
+    port: 3389
+    region: "{{ secondary_cluster_region }}"
+
+- name: Update cluster port and DO NOT WAIT for remove secondary DB cluster from global DB to complete
+  amazon.aws.rds_cluster:
+    db_cluster_identifier: "{{ secondary_cluster_name }}"
+    global_cluster_identifier: "{{ global_cluster_name }}"
+    remove_from_global_db: true
+    state: present
+    port: 3389
+    region: "{{ secondary_cluster_region }}"
+    wait: false
 """
 
 RETURN = r"""
@@ -1102,6 +1145,33 @@ def ensure_present(cluster, parameters, method_name, method_options_name):
     return changed
 
 
+def handle_remove_from_global_db(module, cluster):
+    global_cluster_id = module.params.get("global_cluster_identifier")
+    db_cluster_id = module.params.get("db_cluster_identifier")
+    db_cluster_arn = cluster["DBClusterArn"]
+
+    if module.check_mode:
+        return True
+
+    try:
+        client.remove_from_global_cluster(DbClusterIdentifier=db_cluster_arn, GlobalClusterIdentifier=global_cluster_id)
+    except (botocore.exceptions.BotoCoreError, botocore.exceptions.ClientError) as e:
+        module.fail_json_aws(
+            e, msg=f"Failed to remove cluster {db_cluster_id} from global DB cluster {global_cluster_id}."
+        )
+
+    # for replica cluster - wait for cluster to change status from 'available' to 'promoting'
+    # only replica/secondary clusters have "GlobalWriteForwardingStatus" field
+    if "GlobalWriteForwardingStatus" in cluster:
+        wait_for_cluster_status(client, module, db_cluster_id, "db_cluster_promoting")
+
+    # if wait=true, wait for db cluster remove from global db operation to complete
+    if module.params.get("wait"):
+        wait_for_cluster_status(client, module, db_cluster_id, "cluster_available")
+
+    return True
+
+
 def main():
     global module
     global client
@@ -1154,6 +1224,7 @@ def main():
         port=dict(type="int"),
         preferred_backup_window=dict(aliases=["backup_window"]),
         preferred_maintenance_window=dict(aliases=["maintenance_window"]),
+        remove_from_global_db=dict(type="bool"),
         replication_source_identifier=dict(aliases=["replication_src_id"]),
         restore_to_time=dict(),
         restore_type=dict(choices=["full-copy", "copy-on-write"]),
@@ -1190,6 +1261,7 @@ def main():
         required_if=[
             ["creation_source", "snapshot", ["snapshot_identifier", "engine"]],
             ["creation_source", "s3", required_by_s3_creation_source],
+            ["remove_from_global_db", True, ["global_cluster_identifier", "db_cluster_identifier"]],
         ],
         mutually_exclusive=[
             ["s3_bucket_name", "source_db_cluster_identifier", "snapshot_identifier"],
@@ -1245,12 +1317,17 @@ def main():
             msg="skip_final_snapshot is False but all of the following are missing: final_snapshot_identifier"
         )
 
-    parameters = arg_spec_to_rds_params(dict((k, module.params[k]) for k in module.params if k in parameter_options))
     changed = False
+
+    parameters = arg_spec_to_rds_params(dict((k, module.params[k]) for k in module.params if k in parameter_options))
     method_name, method_options_name = get_rds_method_attribute_name(cluster)
 
     if method_name:
         if method_name == "delete_db_cluster":
+            if cluster and module.params.get("remove_from_global_db"):
+                if cluster["Engine"] in ["aurora", "aurora-mysql", "aurora-postgresql"]:
+                    changed = handle_remove_from_global_db(module, cluster)
+
             call_method(client, module, method_name, eval(method_options_name)(parameters))
             changed = True
         else:
@@ -1261,6 +1338,12 @@ def main():
     else:
         cluster_id = module.params["db_cluster_identifier"]
 
+    if cluster_id and get_cluster(cluster_id) and module.params.get("remove_from_global_db"):
+        if cluster["Engine"] in ["aurora", "aurora-mysql", "aurora-postgresql"]:
+            if changed:
+                wait_for_cluster_status(client, module, cluster_id, "cluster_available")
+        changed |= handle_remove_from_global_db(module, cluster)
+
     result = camel_dict_to_snake_dict(get_cluster(cluster_id))
 
     if result:

tests/integration/requirements.yml

@@ -2,3 +2,4 @@
 collections:
 - ansible.windows
 - ansible.utils  # ipv6 filter
+- amazon.cloud # used by integration tests - rds_cluster_modify

tests/integration/targets/rds_cluster_modify/defaults/main.yml

@@ -12,3 +12,20 @@ new_cluster_id: ansible-test-cluster-{{ tiny_prefix }}-new
 new_port: 1155
 new_password: test-rds_password-new
 new_db_parameter_group_name: ansible-test-db-parameter-group-{{ tiny_prefix }}-new
+
+test_engine: aurora-mysql
+test_engine_version: 8.0
+test_instance_class: db.r5.large
+
+# Global cluster parameters ================================
+test_global_cluster_name: ansible-test-global-{{ tiny_prefix }}
+
+# Primary cluster parameters ================================
+test_primary_cluster_name: ansible-test-primary-{{ tiny_prefix }}
+test_primary_cluster_region: us-west-2
+test_primary_cluster_instance_name: ansible-test-instance-primary-{{ tiny_prefix }}
+
+# Replica cluster parameters ================================
+test_replica_cluster_name: ansible-test-replica-{{ tiny_prefix }}
+test_replica_cluster_region: eu-north-1
+test_replica_cluster_instance_name: ansible-test-instance-replica-{{ tiny_prefix }}

tests/integration/targets/rds_cluster_modify/tasks/main.yaml

@@ -5,6 +5,12 @@
       secret_key: '{{ aws_secret_key }}'
       session_token: '{{ security_token | default(omit) }}'
   block:
+
+  # Disabled: Below tests require use of more than 1 region, not supported by CI at the moment
+  # Tests have been ran, tested, and verified locally on us-west-2 (primary), eu-north-1 (replica)
+  # - name: Run tests for testing remove cluster from global db
+  #   import_tasks: remove_from_global_db.yaml
+
   - name: Ensure the resource doesn't exist
     rds_cluster:
       id: '{{ cluster_id }}'
@@ -177,7 +183,7 @@
 
   - name: Create DB cluster parameter group if not exists
     command: aws rds create-db-cluster-parameter-group --db-cluster-parameter-group-name
-      {{ new_db_parameter_group_name }} --db-parameter-group-family aurora-mysql5.7 --description
+      {{ new_db_parameter_group_name }} --db-parameter-group-family aurora-mysql8.0 --description
       "Test DB cluster parameter group"
     environment:
       AWS_ACCESS_KEY_ID: '{{ aws_access_key }}'