dongjoon-hyun commented on code in PR #57355:
URL: https://github.com/apache/spark/pull/57355#discussion_r4081857954
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/exchange/ShuffleExchangeExec.scala:
##########
@@ -360,6 +370,7 @@ object ShuffleExchangeExec {
newPartitioning: Partitioning,
serializer: Serializer,
writeMetrics: Map[String, SQLMetric],
+ customWriteMetrics: Map[String, SQLMetric],
Review Comment:
This is still binary-incompatible. In `branch-4.3` and `master`, the
signature is `(rdd, outputAttributes, newPartitioning, serializer,
writeMetrics: Map, pipelined: Boolean = false)`. Since Scala compiles an
existing 5-arg call site into a call to the 6-arg JVM method plus
`prepareShuffleDependency$default$6`, downstream jars compiled against 4.3 will
hit `NoSuchMethodError` because `(..., Map, boolean)` no longer exists. The new
5-arg overload does not help for this. In addition, a caller passing
`pipelined` positionally, `(..., writeMetrics, true)`, fails at source level
too.
Could you keep the existing `(..., writeMetrics: Map[String, SQLMetric],
pipelined: Boolean = false)` signature as-is (delegating to the new method),
and make the new method with `customWriteMetrics` have no default arguments?
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
