Github user walterddr commented on a diff in the pull request: https://github.com/apache/flink/pull/5555#discussion_r183222573 --- Diff: flink-libraries/flink-table/src/test/scala/org/apache/flink/table/runtime/stream/sql/OverWindowITCase.scala --- @@ -50,6 +51,96 @@ class OverWindowITCase extends StreamingWithStateTestBase { (8L, 8, "Hello World"), (20L, 20, "Hello World")) + @Test + def testProcTimeDistinctBoundedPartitionedRowsOver(): Unit = { --- End diff -- Both case went through: `AggregateExpandDistinctAggregatesRule`, thus produce plans that breaks the distinct aggregate into a `GROUP BY` over distinct key, and aggregate without distinct. In the 2nd case, `LogicalWindowAggregateRule` did not applied due to `distinct` keyword. They should both be fixed in FLINK-8690
---