HeartSaVioR commented on a change in pull request #26162: [SPARK-29438][SS] Use
partition ID of StateStoreAwareZipPartitionsRDD for determining partition ID of
state store in stream-stream join
URL: https://github.com/apache/spark/pull/26162#discussion_r343964584
##########
File path:
sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/StreamingSymmetricHashJoinHelper.scala
##########
@@ -203,17 +202,18 @@ object StreamingSymmetricHashJoinHelper extends Logging {
/**
* A custom RDD that allows partitions to be "zipped" together, while
ensuring the tasks'
* preferred location is based on which executors have the required join
state stores already
- * loaded. This is class is a modified version of [[ZippedPartitionsRDD2]].
+ * loaded. This class is a variant of [[ZippedPartitionsRDD2]] which only
changes signature
+ * of `f`.
*/
class StateStoreAwareZipPartitionsRDD[A: ClassTag, B: ClassTag, V: ClassTag](
sc: SparkContext,
- f: (Iterator[A], Iterator[B]) => Iterator[V],
- rdd1: RDD[A],
- rdd2: RDD[B],
+ var f: (Int, Iterator[A], Int, Iterator[B]) => Iterator[V],
Review comment:
This is to making them null in `clearDependencies` which was done in
`ZippedPartitionsRDD2`. Now we can't extend ZippedPartitionsRDD2 so we have to
copy it in here.
----------------------------------------------------------------
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]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
