rdblue commented on a change in pull request #28523:
URL: https://github.com/apache/spark/pull/28523#discussion_r424711806
##########
File path:
sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/StreamExecution.scala
##########
@@ -629,14 +630,9 @@ abstract class StreamExecution(
writeBuilder.asInstanceOf[SupportsTruncate].truncate().buildForStreaming()
case Update =>
- // Although no v2 sinks really support Update mode now, but during
tests we do want them
- // to pretend to support Update mode, and treat Update mode same as
Append mode.
- if (Utils.isTesting) {
Review comment:
Is this what prevented tests from catching that v2 sinks didn't support
update?
I find it _very_ concerning that the default implementations were changed to
v2 when sinks don't actually support update.
How do we plan to verify that the v2 path works, since we don't have
reliable test coverage? Maybe we should move back to v1 after all.
----------------------------------------------------------------
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:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
