You are viewing a plain text version of this content. The canonical link for it is here.

Posted to github@beam.apache.org by GitBox <gi...@apache.org> on 2020/07/16 00:51:37 UTC

[GitHub] [beam] ihji commented on a change in pull request #12164: [BEAM-10397] add missing environment in windowing strategy for Dataflow

ihji commented on a change in pull request #12164:
URL: https://github.com/apache/beam/pull/12164#discussion_r455446568



##########
File path: sdks/python/apache_beam/runners/dataflow/dataflow_runner.py
##########
@@ -864,7 +867,7 @@ def run_GroupByKey(self, transform_node, options):
     windowing = transform_node.transform.get_windowing(transform_node.inputs)
     step.add_property(
         PropertyNames.SERIALIZED_FN,
-        self.serialize_windowing_strategy(windowing))
+        self.serialize_windowing_strategy(windowing, self._default_environment))

Review comment:
       @robertwb The only problematic scenario is when `WindowingStrategy` comes from different SDK and the proto doesn't have a value in `environment_id` field. In that case, the host SDK will populate the field with its default environment. But I think it shouldn't happen since it's expansion service's responsibility to pass a semantically correct proto.




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org