[AIRFLOW-2797] Create Google Dataproc cluster with custom image (apache#3871)

Author: exploy

airflow/contrib/operators/dataproc_operator.py

@@ -67,7 +67,10 @@ class DataprocClusterCreateOperator(BaseOperator):
         to add to all instances
     :type metadata: dict
     :param image_version: the version of software inside the Dataproc cluster
-    :type image_version: string
+    :type image_version: str
+    :param custom_image: custom Dataproc image for more info see
+        https://cloud.google.com/dataproc/docs/guides/dataproc-images
+    :type: custom_image: str
     :param properties: dict of properties to set on
         config files (e.g. spark-defaults.conf), see
         https://cloud.google.com/dataproc/docs/reference/rest/v1/ \
@@ -138,6 +141,7 @@ def __init__(self,
                  init_actions_uris=None,
                  init_action_timeout="10m",
                  metadata=None,
+                 custom_image=None,
                  image_version=None,
                  properties=None,
                  master_machine_type='n1-standard-4',
@@ -168,6 +172,7 @@ def __init__(self,
         self.init_actions_uris = init_actions_uris
         self.init_action_timeout = init_action_timeout
         self.metadata = metadata
+        self.custom_image = custom_image
         self.image_version = image_version
         self.properties = properties
         self.master_machine_type = master_machine_type
@@ -187,6 +192,9 @@ def __init__(self,
         self.auto_delete_time = auto_delete_time
         self.auto_delete_ttl = auto_delete_ttl
 
+        assert not (self.custom_image and self.image_version), \
+            "custom_image and image_version can't be both set"
+
     def _get_cluster_list_for_project(self, service):
         result = service.projects().regions().clusters().list(
             projectId=self.project_id,
@@ -321,6 +329,12 @@ def _build_cluster_data(self):
             cluster_data['config']['gceClusterConfig']['tags'] = self.tags
         if self.image_version:
             cluster_data['config']['softwareConfig']['imageVersion'] = self.image_version
+        elif self.custom_image:
+            custom_image_url = 'https://www.googleapis.com/compute/beta/projects/' \
+                               '{}/global/images/{}'.format(self.project_id,
+                                                            self.custom_image)
+            cluster_data['config']['masterConfig']['imageUri'] = custom_image_url
+            cluster_data['config']['workerConfig']['imageUri'] = custom_image_url
         if self.properties:
             cluster_data['config']['softwareConfig']['properties'] = self.properties
         if self.idle_delete_ttl:

tests/contrib/operators/test_dataproc_operator.py

@@ -59,6 +59,7 @@
 TAGS = ['tag1', 'tag2']
 STORAGE_BUCKET = 'gs://airflow-test-bucket/'
 IMAGE_VERSION = '1.1'
+CUSTOM_IMAGE = 'test-custom-image'
 MASTER_MACHINE_TYPE = 'n1-standard-2'
 MASTER_DISK_SIZE = 100
 WORKER_MACHINE_TYPE = 'n1-standard-2'
@@ -258,6 +259,39 @@ def test_build_cluster_data_with_autoDeleteTime_and_autoDeleteTtl(self):
         self.assertEqual(cluster_data['config']['lifecycleConfig']['autoDeleteTime'],
                          "2017-06-07T00:00:00.000000Z")
 
+    def test_init_with_image_version_and_custom_image_both_set(self):
+        with self.assertRaises(AssertionError):
+            DataprocClusterCreateOperator(
+                task_id=TASK_ID,
+                cluster_name=CLUSTER_NAME,
+                project_id=PROJECT_ID,
+                num_workers=NUM_WORKERS,
+                zone=ZONE,
+                dag=self.dag,
+                image_version=IMAGE_VERSION,
+                custom_image=CUSTOM_IMAGE
+            )
+
+    def test_init_with_custom_image(self):
+        dataproc_operator = DataprocClusterCreateOperator(
+            task_id=TASK_ID,
+            cluster_name=CLUSTER_NAME,
+            project_id=PROJECT_ID,
+            num_workers=NUM_WORKERS,
+            zone=ZONE,
+            dag=self.dag,
+            custom_image=CUSTOM_IMAGE
+        )
+
+        cluster_data = dataproc_operator._build_cluster_data()
+        expected_custom_image_url = \
+            'https://www.googleapis.com/compute/beta/projects/' \
+            '{}/global/images/{}'.format(PROJECT_ID, CUSTOM_IMAGE)
+        self.assertEqual(cluster_data['config']['masterConfig']['imageUri'],
+                         expected_custom_image_url)
+        self.assertEqual(cluster_data['config']['workerConfig']['imageUri'],
+                         expected_custom_image_url)
+
     def test_cluster_name_log_no_sub(self):
         with patch('airflow.contrib.operators.dataproc_operator.DataProcHook') \
                 as mock_hook: