Kenn Knowles
Kenn Knowles
Based on this getting pushed from release to release, it is clearly not a true release-blocker.
Yea, let's work out how to get top-level signal.
Could make sense to find a way to get separate top-level signal for Python versions, assuming we can use software engineering to share everything necessary so they don't get out...
I looked at a couple flakes and could not discern if they represented anything that should be release blocking, so I am moving this to the next release milestone.
@damccorm this one is perma-red. Should we disable or is it obsolete or...?
Notes on failures: - ValidatesRunner Dataflow: failures are all side input tests, leftover from `master` breakage; we already know this was green from the first time - ValidatesRunner Dataflow JavaVersions:...
So in conclusion, my analysis is this: Non-portable: - DirectRunner works - Dataflow works - Samza works - Flink works - Spark StructuredStreaming works - Spark batch has some issue...
How about we delete this suite? I don't think we get any benefit from running it.
For the reviewer, summary so far: - **Roll forward**: I am rolling forward just the composite transform and tests, which are pretty trivial, also the same as Reshuffle, and already...