damccorm commented on code in PR #32325:
URL: https://github.com/apache/beam/pull/32325#discussion_r1733586276
##########
sdks/python/apache_beam/runners/direct/direct_runner.py:
##########
@@ -516,7 +516,8 @@ def visit_transform(self, applied_ptransform):
if isinstance(applied_ptransform.transform, ExternalTransform):
raise RuntimeError(
"Streaming Python direct runner "
- "does not support cross-language pipelines.")
+ "does not support cross-language pipelines."
+ "Please use other runners such as DataflowRunner.")
Review Comment:
> I would love for Python users to try Prism out before we make the
recommendation though. All I do is run the test suites. I don't try running
proper pipelines at this stage.
IMO this is actually the perfect place to get users to try prism. What
they've just tried didn't work, so at worst we're providing an equal experience
(and hopefully we're providing a better one).
> If this were to say "...such as FlinkRunner, DataflowRunner, or
PrismRunner", that would avoid the conflict problem though. (IIUC, Python will
download and launch a Flink instance, just as it would for Prism).
I think this is a reasonable option, though I'd still vote we just use this
as an opportunity to guide towards prism. I think that is the best way to start
exposing gaps that may exist, and this seems like a low risk environment to do
so
--
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]
