Github user JoshRosen commented on a diff in the pull request:
https://github.com/apache/spark/pull/9383#discussion_r43786160
--- Diff:
sql/core/src/test/scala/org/apache/spark/sql/execution/UnsafeFixedWidthAggregationMapSuite.scala
---
@@ -291,28 +278,27 @@ class UnsafeFixedWidthAggregationMapSuite
val sorter = map.destructAndCreateExternalSorter()
// Add more keys to the sorter and make sure the results come out
sorted.
+ map = createMap()
(1 to 4096).foreach { i =>
- sorter.insertKV(UnsafeRow.createFromByteArray(0, 0),
UnsafeRow.createFromByteArray(0, 0))
+
map.getAggregationBufferFromUnsafeRow(UnsafeRow.createFromByteArray(0, 0))
if ((i % 100) == 0) {
- memoryManager.markExecutionAsOutOfMemoryOnce()
- sorter.closeCurrentPage()
+ val sorter2 = map.destructAndCreateExternalSorter()
+ sorter.merge(sorter2)
+ map = createMap()
}
}
+ val sorter2 = map.destructAndCreateExternalSorter()
+ sorter.merge(sorter2)
var count = 0
val iter = sorter.sortedIterator()
while (iter.next()) {
- // At here, we also test if copy is correct.
- iter.getKey.copy()
- iter.getValue.copy()
count += 1
}
- // 1 record was from the map and 4096 records were explicitly inserted.
- assert(count === 4097)
-
- map.free()
+ // 1 record in one map.
+ assert(count === 42)
--- End diff --
Why this change? Shouldn't the total number of records have been unchanged
by this refactoring?
---
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]
