[SPARK-24935][SQL][2.3] fix Hive UDAF with two aggregation buffers #24539
Conversation
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
## What changes were proposed in this pull request? Hive UDAF knows the aggregation mode when creating the aggregation buffer, so that it can create different buffers for different inputs: the original data or the aggregation buffer. Please see an example in the [sketches library](https://github.com/DataSketches/sketches-hive/blob/7f9e76e9e03807277146291beb2c7bec40e8672b/src/main/java/com/yahoo/sketches/hive/cpc/DataToSketchUDAF.java#L107). However, the Hive UDAF adapter in Spark always creates the buffer with partial1 mode, which can only deal with one input: the original data. This PR fixes it. All credits go to pgandhi999 , who investigate the problem and study the Hive UDAF behaviors, and write the tests. close apache#23778 ## How was this patch tested? a new test Closes apache#24144 from cloud-fan/hive. Lead-authored-by: pgandhi <[email protected]> Co-authored-by: Wenchen Fan <[email protected]> Signed-off-by: gatorsmile <[email protected]>
…H in Hive UDAF adapter ## What changes were proposed in this pull request? This is a followup of apache#24144 . apache#24144 missed one case: when hash aggregate fallback to sort aggregate, the life cycle of UDAF is: INIT -> UPDATE -> MERGE -> FINISH. However, not all Hive UDAF can support it. Hive UDAF knows the aggregation mode when creating the aggregation buffer, so that it can create different buffers for different inputs: the original data or the aggregation buffer. Please see an example in the [sketches library](https://github.com/DataSketches/sketches-hive/blob/7f9e76e9e03807277146291beb2c7bec40e8672b/src/main/java/com/yahoo/sketches/hive/cpc/DataToSketchUDAF.java#L107). The buffer for UPDATE may not support MERGE. This PR updates the Hive UDAF adapter in Spark to support INIT -> UPDATE -> MERGE -> FINISH, by turning it to INIT -> UPDATE -> FINISH + IINIT -> MERGE -> FINISH. ## How was this patch tested? a new test case Closes apache#24459 from cloud-fan/hive-udaf. Authored-by: Wenchen Fan <[email protected]> Signed-off-by: Wenchen Fan <[email protected]>
Contributor
Author
|
cc @gatorsmile |
|
+1 |
|
Test build #105159 has finished for PR 24539 at commit
|
dongjoon-hyun
approved these changes
May 6, 2019
Member
dongjoon-hyun
left a comment
dongjoon-hyun
left a comment
There was a problem hiding this comment.
+1, LGTM. This is a clean cherry-pick of both patches.
Merged to branch-2.3
dongjoon-hyun
pushed a commit
that referenced
this pull request
May 6, 2019
## What changes were proposed in this pull request? backport #24144 and #24459 to 2.3. ## How was this patch tested? existing tests Closes #24539 from cloud-fan/backport. Lead-authored-by: pgandhi <[email protected]> Co-authored-by: Wenchen Fan <[email protected]> Signed-off-by: Dongjoon Hyun <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
What changes were proposed in this pull request?
backport #24144 and #24459 to 2.3.
How was this patch tested?
existing tests