andygrove commented on code in PR #4791:
URL: https://github.com/apache/datafusion-comet/pull/4791#discussion_r3532497636
##########
spark/src/test/scala/org/apache/comet/exec/CometAggregateSuite.scala:
##########
@@ -45,8 +45,17 @@ class CometAggregateSuite extends CometTestBase with
AdaptiveSparkPlanHelper {
// Several aggregate tests exercise overflow behavior expected to wrap
around silently;
// ANSI-mode variants opt in to ANSI explicitly via withSQLConf.
+ //
+ // For the CUBE(9) offset-overflow repro at the bottom of the suite we MUST
disable
+ // off-heap (CometTestBase defaults to off-heap + 2 GiB cap, which spills
long before
+ // i32::MAX) and force the on-heap memory pool to `unbounded`. Both knobs
are read
+ // from SparkConf at SparkContext init (CometExecIterator.scala:295/310),
NOT from
+ // per-query SQLConf, so they must live here.
override protected def sparkConf: SparkConf =
- super.sparkConf.set(SQLConf.ANSI_ENABLED.key, "false")
+ super.sparkConf
+ .set(SQLConf.ANSI_ENABLED.key, "false")
+ .set("spark.memory.offHeap.enabled", "false")
Review Comment:
These changes seem specific to the ignored test, but will impact all of the
existing tests?
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
