redsk commented on a change in pull request #26153: [SPARK-29500][SQL][SS]
Support partition column when writing to Kafka
URL: https://github.com/apache/spark/pull/26153#discussion_r336458516
##########
File path:
external/kafka-0-10-sql/src/test/scala/org/apache/spark/sql/kafka010/KafkaSinkSuite.scala
##########
@@ -404,20 +420,85 @@ abstract class KafkaSinkBatchSuiteBase extends
KafkaSinkSuiteBase {
.save()
checkAnswer(
createKafkaReader(topic, includeHeaders = true).selectExpr(
- "CAST(value as STRING) value", "headers"
+ "CAST(value as STRING) value", "headers", "partition"
),
- Row("1", Seq(Row("a", "b".getBytes(UTF_8)))) ::
- Row("2", Seq(Row("c", "d".getBytes(UTF_8)), Row("e",
"f".getBytes(UTF_8)))) ::
- Row("3", Seq(Row("g", "h".getBytes(UTF_8)), Row("g",
"i".getBytes(UTF_8)))) ::
- Row("4", null) ::
+ Row("1", Seq(Row("a", "b".getBytes(UTF_8))), 0) ::
+ Row("2", Seq(Row("c", "d".getBytes(UTF_8)), Row("e",
"f".getBytes(UTF_8))), 1) ::
+ Row("3", Seq(Row("g", "h".getBytes(UTF_8)), Row("g",
"i".getBytes(UTF_8))), 2) ::
+ Row("4", null, 3) ::
Row("5", Seq(
Row("j", "k".getBytes(UTF_8)),
Row("j", "l".getBytes(UTF_8)),
- Row("m", "n".getBytes(UTF_8)))) ::
+ Row("m", "n".getBytes(UTF_8))), 0) ::
Nil
)
}
+ test("batch - partition column vs default Kafka partitioner") {
Review comment:
> I think `test("batch - write to kafka")` should be the simplest test and
shouldn't contain neither headers nor partitions (I know for headers it's a bit
late).
I can revert it to the previous state (with headers). Is that ok?
----------------------------------------------------------------
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]
