Github user tdas commented on a diff in the pull request:
https://github.com/apache/spark/pull/15801#discussion_r87315895
--- Diff:
sql/core/src/test/scala/org/apache/spark/sql/streaming/test/DataStreamReaderWriterSuite.scala
---
@@ -467,4 +468,54 @@ class DataStreamReaderWriterSuite extends StreamTest
with BeforeAndAfter {
val sq = df.writeStream.format("console").start()
sq.stop()
}
+
+ test("MemorySink can recover from a checkpoint in Complete Mode") {
+ import testImplicits._
+ val ms = new MemoryStream[Int](0, sqlContext)
+ val df = ms.toDF().toDF("a")
+ val checkpointLoc = newMetadataDir
+ val checkpointDir = new File(checkpointLoc, "offsets")
+ checkpointDir.mkdirs()
+ assert(checkpointDir.exists())
+ def startStream: StreamingQuery = {
+ df.groupBy("a")
+ .count()
+ .writeStream
+ .format("memory")
+ .queryName("test")
+ .option("checkpointLocation", checkpointLoc)
+ .outputMode("complete")
+ .start()
+ }
+ // no exception here
+ val q = startStream
+ ms.addData(0, 1)
+ q.processAllAvailable()
+ q.stop()
+
+ checkAnswer(
+ spark.table("test"),
+ Seq(Row(0, 1), Row(1, 1))
+ )
+
+ val q2 = startStream
+ q2.processAllAvailable()
+ checkAnswer(
+ spark.table("test"),
+ Seq(Row(0, 1), Row(1, 1))
+ )
--- End diff --
its not clear if this table was regenerated by the query, or just the same
table that existed before when the query was stopped earlier. so it would be
good it if before restarting you remove the table, or rewrite the table to
something different (and verify that), so that we are sure that the table is
actually being regenerated.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
