Repository navigation
test: skip two Spark 3.4 tests that cannot pass there - #7423
Merged
Merged
Conversation
The public foreachPartition observation tests rely on a listener callback that Spark 3.4 does not fire, because Dataset.foreachPartition runs unnamed there. The Iceberg SPARK-34637 DPP test fails on 3.4 with a native planning error tracked in apache#7422.
github-actions
Bot
added
enhancement
mbutrovich
approved these changes
Nov 2, 2026
Contributor
There was a problem hiding this comment.
Usually I don't love a workaround that just disables a failing test for a Spark version, but it's Spark 3.4...
Thanks @andygrove!
Contributor
|
The fix #7424 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Which issue does this PR close?
Related to #7422. Does not close it: the Iceberg failure it describes is still open.
Rationale for this change
The Spark 3.4 jobs fail on two tests that were never run against 3.4 when they merged. They only surfaced on #7402, which carried
run-all-spark-profiles.What changes are included in this PR?
CometCollectMetricsSuite: the two "public foreachPartition preserves observation row demand" tests now require Spark 3.5+. Spark 3.4'sDataset.foreachPartitioncallswithNewRDDExecutionIdwithout a name, so noQueryExecutionListenercallback reports the plan the test asserts on.CometIcebergNativeSuite: "AQE DPP - V2 broadcast query stage creation order (SPARK-34637)" now requires Spark 3.5+. On 3.4 it fails deterministically withColumn index 1 is out of bound. Schema: .The root cause is a real bug tracked in Spark 3.4: native Iceberg scan with AQE DPP fails with "Column index 1 is out of bound" when the build side needs an inserted exchange #7422, so the test should be re-enabled when that is fixed. The now-deadisSpark35Pluscheck inside the test is removed.How are these changes tested?
Ran each test locally with
-Pspark-3.4: all three report as canceled, and before the change the Iceberg test failed with the error above. The Spark 3.5+ behavior is unchanged.