Merge pull request #1211 from shreyabiradar07/test_multiple_import

Test multiple metadata import without datasource connection

tests/scripts/helpers/utils.py

@@ -55,6 +55,7 @@
 LIST_METADATA_ERROR_MSG = ("Metadata for a given datasource - %s, cluster name - %s, namespace - %s "
                            "either does not exist or is not valid")
 LIST_METADATA_DATASOURCE_NAME_CLUSTER_NAME_ERROR_MSG = "Metadata for a given datasource name - %s, cluster_name - %s either does not exist or is not valid"
+IMPORT_METADATA_DATASOURCE_CONNECTION_FAILURE_MSG = "Metadata cannot be imported, datasource connection refused or timed out"
 
 # Kruize Recommendations Notification codes
 NOTIFICATION_CODE_FOR_RECOMMENDATIONS_AVAILABLE = "111000"
@@ -935,6 +936,89 @@ def delete_namespace(namespace_name):
         else:
             print(f"Exception deleting namespace: {e}")
 
+
+def scale_deployment(namespace, deployment_name, replicas):
+    """
+    Scale a Kubernetes Deployment to the desired number of replicas.
+
+    This function scales a specified Deployment in a given namespace to a specified number
+    of replicas using the Kubernetes Python client library. It achieves this by creating
+    a Scale object and using the AppsV1Api to update the Deployment's scale.
+
+    Args:
+    - namespace (str): The namespace of the Deployment.
+    - deployment_name (str): The name of the Deployment.
+    - replicas (int): The desired number of replicas.
+
+    Returns:
+    None
+    """
+    config.load_kube_config()  # Load kube config from default location
+
+    # Create an API client
+    apps_v1 = client.AppsV1Api()
+
+    # Define the scale object
+    scale = client.V1Scale(
+        api_version='autoscaling/v1',
+        kind='Scale',
+        metadata=client.V1ObjectMeta(name=deployment_name, namespace=namespace),
+        spec=client.V1ScaleSpec(replicas=replicas)
+    )
+
+    # Scale the deployment
+    try:
+        response = apps_v1.replace_namespaced_deployment_scale(
+            name=deployment_name,
+            namespace=namespace,
+            body=scale
+        )
+        print(f"Deployment {deployment_name} scaled to {replicas} replicas successfully.")
+    except client.exceptions.ApiException as e:
+        print(f"Error scaling deployment {deployment_name}: {e}")
+
+
+def scale_statefulset(namespace, statefulset_name, replicas):
+    """
+    Scale a Kubernetes Statefulset to the desired number of replicas.
+
+    This function scales a specified Statefulset in a given namespace to a specified number
+    of replicas using the Kubernetes Python client library. It achieves this by creating
+    a Scale object and using the AppsV1Api to update the Statefulset's scale.
+
+    Args:
+    - namespace (str): The namespace of the Deployment.
+    - statefulset_name (str): The name of the Statefulset.
+    - replicas (int): The desired number of replicas.
+
+    Returns:
+    None
+    """
+    config.load_kube_config()  # Load kube config from default location
+
+    # Create an API client
+    apps_v1 = client.AppsV1Api()
+
+    # Define the scale object
+    scale = client.V1Scale(
+        api_version='autoscaling/v1',
+        kind='Scale',
+        metadata=client.V1ObjectMeta(name=statefulset_name, namespace=namespace),
+        spec=client.V1ScaleSpec(replicas=replicas)
+    )
+
+    # Scale the statefulset
+    try:
+        response = apps_v1.replace_namespaced_stateful_set_scale(
+            name=statefulset_name,
+            namespace=namespace,
+            body=scale
+        )
+        print(f"StatefulSet {statefulset_name} scaled to {replicas} replicas successfully.")
+    except client.exceptions.ApiException as e:
+        print(f"Error scaling statefulset {statefulset_name}: {e}")
+
+
 # validate duration_in_hours decimal precision
 def validate_duration_in_hours_decimal_precision(duration_in_hours):
     """

tests/scripts/local_monitoring_tests/Local_monitoring_tests.md

@@ -20,6 +20,12 @@ Here are the test scenarios:
 - Post the same datasource again
 - Test with invalid values such as blank, null or an invalid value for various keys in the dsmetadata input request json
 - Validate error messages when the mandatory fields are missing
+- Repeated metadata imports for the same datasource by dynamically creating and deleting namespaces between two metadata
+  import actions by validating the newly created namespaces by verifying listing metadata output after the second import
+  metadata invocation
+- Repeated metadata imports without datasource connection by dynamically scaling down prometheus resources to zero replicas
+  and validating the behaviour of import metadata resulting in an error when datasource cannot be connected. Additionally,
+  to verify the output of list metadata returns the metadata from the DB after the second import metadata invocation.
 
 ### **List Metadata API tests**
 

tests/scripts/local_monitoring_tests/rest_apis/test_import_metadata.py

@@ -148,7 +148,8 @@ def test_import_metadata_mandatory_fields(cluster_type, field, expected_status_c
 def test_repeated_metadata_import(cluster_type):
     """
     Test Description: This test validates the response status code of /dsmetadata API by specifying the
-    same datasource name
+    same datasource name by creating and deleting namespaces between two import metadata calls and validating the newly
+    created namespaces by verifying listing metadata output after the second import metadata invocation
     """
     input_json_file = "../json_files/import_metadata.json"
     json_data = json.load(open(input_json_file))
@@ -224,3 +225,87 @@ def test_repeated_metadata_import(cluster_type):
 
     delete_namespace("repeated-metadata-import")
     delete_namespace("local-monitoring-test")
+
+
+@pytest.mark.negative
+def test_repeated_metadata_import_without_datasource_connection(cluster_type):
+    """
+    Test Description: This test validates the response status code of POST /dsmetadata API by specifying the
+    same datasource name with repeated metadata imports by bringing down prometheus server instance to validate the behaviour
+    of import metadata when datasource cannot be connected resulting in an error and additionally verifying list metadata
+    output returns the metadata from the DB after the second import metadata invocation.
+    """
+    input_json_file = "../json_files/import_metadata.json"
+    json_data = json.load(open(input_json_file))
+
+    datasource_name = json_data['datasource_name']
+    print("datasource_name = ", datasource_name)
+
+    form_kruize_url(cluster_type)
+
+    response = delete_metadata(input_json_file)
+    print("delete metadata = ", response.status_code)
+
+    # Import metadata using the specified json
+    response = import_metadata(input_json_file)
+    metadata_json = response.json()
+
+    assert response.status_code == SUCCESS_STATUS_CODE
+
+    # Validate the json against the json schema
+    errorMsg = validate_import_metadata_json(metadata_json, import_metadata_json_schema)
+    assert errorMsg == ""
+
+    json_data = json.load(open(input_json_file))
+    datasource = json_data['datasource_name']
+    # Currently only default cluster_name is supported by kruize
+    cluster_name = "default"
+    response = list_metadata(datasource=datasource, cluster_name=cluster_name)
+
+    list_metadata_json = response.json()
+    assert response.status_code == SUCCESS_200_STATUS_CODE
+
+    if cluster_type == "minikube":
+        namespace = "monitoring"
+    elif cluster_type == "openshift":
+        namespace = "openshift-monitoring"
+
+    # Validate the json values
+    import_metadata_json = read_json_data_from_file(input_json_file)
+    validate_list_metadata_parameters(import_metadata_json, list_metadata_json, cluster_name=cluster_name, namespace=namespace)
+
+
+    # Scaling down prometheus deployment and statefulset to zero replicas to bring down prometheus datasource connection
+    scale_deployment(namespace, "prometheus-operator", 0)
+    scale_deployment(namespace, "prometheus-adapter", 0)
+    scale_statefulset(namespace, "prometheus-k8s", 0)
+    time.sleep(10)
+
+    # Repeated Import metadata using the specified json
+    response = import_metadata(input_json_file)
+    metadata_json = response.json()
+
+    assert response.status_code == ERROR_STATUS_CODE
+    assert metadata_json['message'] == IMPORT_METADATA_DATASOURCE_CONNECTION_FAILURE_MSG
+
+    json_data = json.load(open(input_json_file))
+    datasource = json_data['datasource_name']
+    # Currently only default cluster_name is supported by kruize
+    cluster_name = "default"
+    response = list_metadata(datasource=datasource, cluster_name=cluster_name)
+
+    list_metadata_json = response.json()
+    assert response.status_code == SUCCESS_200_STATUS_CODE
+
+    # Validate the json values
+    import_metadata_json = read_json_data_from_file(input_json_file)
+    validate_list_metadata_parameters(import_metadata_json, list_metadata_json, cluster_name=cluster_name, namespace=namespace)
+
+    #validate namespaces
+    response = delete_metadata(input_json_file)
+    print("delete metadata = ", response.status_code)
+
+    scale_deployment(namespace, "prometheus-operator", 1)
+    scale_deployment(namespace, "prometheus-adapter", 2)
+    scale_statefulset(namespace, "prometheus-k8s", 2)
+    time.sleep(90)