[AIRFLOW-2928] Use uuid4 instead of uuid1 (#3779)

for better randomness.
apache · Aug 29, 2018 · c3939c8 · c3939c8
1 parent 314232c
commit c3939c8
Show file tree

Hide file tree

Showing 7 changed files with 12 additions and 12 deletions.
diff --git a/airflow/contrib/hooks/gcp_dataflow_hook.py b/airflow/contrib/hooks/gcp_dataflow_hook.py
@@ -250,7 +250,7 @@ def _build_dataflow_job_name(task_id, append_job_name=True):
                 'letter and ending with a letter or number '.format(task_id))
 
         if append_job_name:
-            job_name = task_id + "-" + str(uuid.uuid1())[:8]
+            job_name = task_id + "-" + str(uuid.uuid4())[:8]
         else:
             job_name = task_id
 

diff --git a/airflow/contrib/hooks/gcp_dataproc_hook.py b/airflow/contrib/hooks/gcp_dataproc_hook.py
@@ -81,7 +81,7 @@ def get(self):
 
 class _DataProcJobBuilder:
     def __init__(self, project_id, task_id, cluster_name, job_type, properties):
-        name = task_id + "_" + str(uuid.uuid1())[:8]
+        name = task_id + "_" + str(uuid.uuid4())[:8]
         self.job_type = job_type
         self.job = {
             "job": {
@@ -141,7 +141,7 @@ def set_python_main(self, main):
         self.job["job"][self.job_type]["mainPythonFileUri"] = main
 
     def set_job_name(self, name):
-        self.job["job"]["reference"]["jobId"] = name + "_" + str(uuid.uuid1())[:8]
+        self.job["job"]["reference"]["jobId"] = name + "_" + str(uuid.uuid4())[:8]
 
     def build(self):
         return self.job

diff --git a/airflow/contrib/kubernetes/pod_generator.py b/airflow/contrib/kubernetes/pod_generator.py
@@ -149,7 +149,7 @@ def make_pod(self, namespace, image, pod_id, cmds, arguments, labels):
 
         return Pod(
             namespace=namespace,
-            name=pod_id + "-" + str(uuid.uuid1())[:8],
+            name=pod_id + "-" + str(uuid.uuid4())[:8],
             image=image,
             cmds=cmds,
             args=arguments,

diff --git a/airflow/contrib/operators/dataflow_operator.py b/airflow/contrib/operators/dataflow_operator.py
@@ -365,7 +365,7 @@ def google_cloud_to_local(self, file_name):
 
         bucket_id = path_components[0]
         object_id = '/'.join(path_components[1:])
-        local_file = '/tmp/dataflow{}-{}'.format(str(uuid.uuid1())[:8],
+        local_file = '/tmp/dataflow{}-{}'.format(str(uuid.uuid4())[:8],
                                                  path_components[-1])
         file_size = self._gcs_hook.download(bucket_id, object_id, local_file)
 

diff --git a/airflow/contrib/operators/dataproc_operator.py b/airflow/contrib/operators/dataproc_operator.py
@@ -1158,7 +1158,7 @@ class DataProcPySparkOperator(BaseOperator):
     @staticmethod
     def _generate_temp_filename(filename):
         dt = time.strftime('%Y%m%d%H%M%S')
-        return "{}_{}_{}".format(dt, str(uuid.uuid1())[:8], ntpath.basename(filename))
+        return "{}_{}_{}".format(dt, str(uuid.uuid4())[:8], ntpath.basename(filename))
 
     """
     Upload a local file to a Google Cloud Storage bucket
@@ -1312,7 +1312,7 @@ def start(self):
             .instantiate(
                 name=('projects/%s/regions/%s/workflowTemplates/%s' %
                       (self.project_id, self.region, self.template_id)),
-                body={'instanceId': str(uuid.uuid1())})
+                body={'instanceId': str(uuid.uuid4())})
             .execute())
 
 
@@ -1355,6 +1355,6 @@ def start(self):
             self.hook.get_conn().projects().regions().workflowTemplates()
             .instantiateInline(
                 parent='projects/%s/regions/%s' % (self.project_id, self.region),
-                instanceId=str(uuid.uuid1()),
+                instanceId=str(uuid.uuid4()),
                 body=self.template)
             .execute())
diff --git a/airflow/contrib/task_runner/cgroup_task_runner.py b/airflow/contrib/task_runner/cgroup_task_runner.py
@@ -123,7 +123,7 @@ def start(self):
         # Create a unique cgroup name
         cgroup_name = "airflow/{}/{}".format(datetime.datetime.utcnow().
                                              strftime("%Y-%m-%d"),
-                                             str(uuid.uuid1()))
+                                             str(uuid.uuid4()))
 
         self.mem_cgroup_name = "memory/{}".format(cgroup_name)
         self.cpu_cgroup_name = "cpu/{}".format(cgroup_name)

diff --git a/tests/contrib/hooks/test_gcp_dataflow_hook.py b/tests/contrib/hooks/test_gcp_dataflow_hook.py
@@ -79,7 +79,7 @@ def setUp(self):
                         new=mock_init):
             self.dataflow_hook = DataFlowHook(gcp_conn_id='test')
 
-    @mock.patch(DATAFLOW_STRING.format('uuid.uuid1'))
+    @mock.patch(DATAFLOW_STRING.format('uuid.uuid4'))
     @mock.patch(DATAFLOW_STRING.format('_DataflowJob'))
     @mock.patch(DATAFLOW_STRING.format('_Dataflow'))
     @mock.patch(DATAFLOW_STRING.format('DataFlowHook.get_conn'))
@@ -103,7 +103,7 @@ def test_start_python_dataflow(self, mock_conn,
         self.assertListEqual(sorted(mock_dataflow.call_args[0][0]),
                              sorted(EXPECTED_CMD))
 
-    @mock.patch(DATAFLOW_STRING.format('uuid.uuid1'))
+    @mock.patch(DATAFLOW_STRING.format('uuid.uuid4'))
     @mock.patch(DATAFLOW_STRING.format('_DataflowJob'))
     @mock.patch(DATAFLOW_STRING.format('_Dataflow'))
     @mock.patch(DATAFLOW_STRING.format('DataFlowHook.get_conn'))
@@ -127,7 +127,7 @@ def test_start_java_dataflow(self, mock_conn,
         self.assertListEqual(sorted(mock_dataflow.call_args[0][0]),
                              sorted(EXPECTED_CMD))
 
-    @mock.patch(DATAFLOW_STRING.format('uuid.uuid1'))
+    @mock.patch(DATAFLOW_STRING.format('uuid.uuid4'))
     @mock.patch(DATAFLOW_STRING.format('_DataflowJob'))
     @mock.patch(DATAFLOW_STRING.format('_Dataflow'))
     @mock.patch(DATAFLOW_STRING.format('DataFlowHook.get_conn'))