Abacn commented on issue #28715: URL: https://github.com/apache/beam/issues/28715#issuecomment-1745366608
> I believe the reason the pipeline fails on 2.50 is because of this: #28399 > > Effectively, when a pipeline is created/run like the example below, it ends up trying to use the Dataflow Runner V1 which is no longer allowed from Beam SDK 2.50+: > > p = beam.Pipeline(InteractiveRunner()) DataflowRunner().run_pipeline(p, options=options) > > A workaround (until 2.51 is released) is to manually specify the "--experiments=use_runner_v2" in the pipeline options. This is unexpected. In Beam 2.49.0 Python, if it does not explicit "disable_runner_v2" experiment, it should default to runner v2, and need `disable_runner_v2` to run on Dataflow legacy runner; in 2.50.0 Python, it should always on runner v2 without the need of specifying experiment. Let us taking a closer look and please send a customer ticket to Dataflow so we can take a look of your jobId -- 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. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
