junaiddshaukat opened a new pull request, #39847:
URL: https://github.com/apache/beam/pull/39847

   Part of #18479, before the merge to master.
   
   The Python modules ship with the SDK because it is published as one package 
and individual modules cannot be held back. They are already inert without the 
job server jar, which is not published with any release, but the failure was 
unhelpful: `path_to_beam_jar` built a Maven Central URL for an artifact that 
does not exist, so the user got a 404.
   
   `path_to_jar` now looks for a jar built from the source tree and, failing 
that, raises an explicit error: the runner is experimental, it is not part of 
any Beam release, and here is how to build the job server. The build commands 
it prints carry `-Pwith-kafka-streams-runner`; without it they fail with 
project not found, which was true of both the message in this file and the 
generic one in `subprocess_server.py`.
   
   Documentation: the runner page now says experimental **and not production 
ready**, with a new section on what happens if you installed Beam from a 
release; the module, runner class and options class carry the same warning; and 
the capability matrix names it "Kafka Streams (experimental, not released)", 
which it did not before.
   
   `ALL_KNOWN_RUNNERS` is unchanged, as discussed.
   
   Tested: 6 Python tests pass, and the two new ones fail if the Maven fallback 
is restored.


-- 
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]

Reply via email to