Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[HUDI-1196] Update HoodieKey when deduplicating records #2248

Merged
merged 1 commit into from
Dec 1, 2020
Merged

[HUDI-1196] Update HoodieKey when deduplicating records #2248

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 3 additions & 4 deletions ...hudi-spark-client/src/main/java/org/apache/hudi/table/action/commit/SparkWriteHelper.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -59,10 +59,9 @@ public JavaRDD<HoodieRecord<T>> deduplicateRecords(JavaRDD<HoodieRecord<T>> reco
}).reduceByKey((rec1, rec2) -> {
@SuppressWarnings("unchecked")
T reducedData = (T) rec1.getData().preCombine(rec2.getData());
// we cannot allow the user to change the key or partitionPath, since that will affect
// everything
// so pick it from one of the records.
return new HoodieRecord<T>(rec1.getKey(), reducedData);
HoodieKey reducedKey = rec1.getData().equals(reducedData) ? rec1.getKey() : rec2.getKey();
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

so if preCombine() is a merge, then the equals check may not succed?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ya I see your point here. From the existing payload implementations precombine either returned this or other so this seemed fine. But if there was other payloads being used with merging behavior, default HoodieKey would be set to rec2.getKey(). This could be solved by regenerating key here, but there would be performance cost

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

let's roll with this for now. If you don't mind, please raise a JIRA for the follow on.


return new HoodieRecord<T>(reducedKey, reducedData);
vinothchandar marked this conversation as resolved.
Show resolved Hide resolved
}, parallelism).map(Tuple2::_2);
}

Expand Down
1 change: 1 addition & 0 deletions ...ark-client/src/test/java/org/apache/hudi/client/TestHoodieClientOnCopyOnWriteStorage.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -241,6 +241,7 @@ private void testDeduplication(
when(index.isGlobal()).thenReturn(true);
List<HoodieRecord<RawTripTestPayload>> dedupedRecs = SparkWriteHelper.newInstance().deduplicateRecords(records, index, 1).collect();
assertEquals(1, dedupedRecs.size());
assertEquals(dedupedRecs.get(0).getPartitionPath(), recordThree.getPartitionPath());
assertNodupesWithinPartition(dedupedRecs);

// non-Global dedup should be done based on both recordKey and partitionPath
Expand Down