c21 opened a new pull request #29216:
URL: https://github.com/apache/spark/pull/29216
<!--
Thanks for sending a pull request! Here are some tips for you:
1. If this is your first time, please read our contributor guidelines:
https://spark.apache.org/contributing.html
2. Ensure you have added or run the appropriate tests for your PR:
https://spark.apache.org/developer-tools.html
3. If the PR is unfinished, add '[WIP]' in your PR title, e.g.,
'[WIP][SPARK-XXXX] Your PR title ...'.
4. Be sure to keep the PR description updated to reflect all changes.
5. Please write your PR title to summarize what this PR proposes.
6. If possible, provide a concise example to reproduce the issue for a
faster review.
7. If you want to add a new configuration, please read the guideline first
for naming configurations in
'core/src/main/scala/org/apache/spark/internal/config/ConfigEntry.scala'.
-->
### What changes were proposed in this pull request?
<!--
Please clarify what changes you are proposing. The purpose of this section
is to outline the changes and how this PR fixes the issue.
If possible, please consider writing useful notes for better and faster
reviews in your PR. See the examples below.
1. If you refactor some codes with changing classes, showing the class
hierarchy will help reviewers.
2. If you fix some SQL features, you can provide some references of other
DBMSes.
3. If there is design documentation, please add the link.
4. If there is a discussion in the mailing list, please add the link.
-->
`HashRelation` has two separate code paths for unique key look up and
non-unique key look up E.g. in its subclass
[`UnsafeHashedRelation`](https://github.com/apache/spark/blob/master/sql/core/src/main/scala/org/apache/spark/sql/execution/joins/HashedRelation.scala#L144-L177),
unique key look up is more efficient as it does not have e.g. extra
`Iterator[UnsafeRow].hasNext()/next()` overhead per row.
`BroadcastHashJoinExec` has handled unique key vs non-unique key separately
in [code-gen
path](https://github.com/apache/spark/blob/master/sql/core/src/main/scala/org/apache/spark/sql/execution/joins/BroadcastHashJoinExec.scala#L289-L321).
But the non-codegen path for broadcast hash join and shuffled hash join do not
separate it yet, so adding the support here.
### Why are the changes needed?
<!--
Please clarify why the changes are needed. For instance,
1. If you propose a new API, clarify the use case for a new API.
2. If you fix a bug, you can clarify why it is a bug.
-->
Shuffled hash join and non-codegen broadcast hash join still rely on this
code path for execution. So this PR will help save CPU for executing this two
type of join. Adding codegen for shuffled hash join would be a different topic
and I will add it in https://issues.apache.org/jira/browse/SPARK-32421 .
Ran the same query as
[`JoinBenchmark`](https://github.com/apache/spark/blob/master/sql/core/src/test/scala/org/apache/spark/sql/execution/benchmark/JoinBenchmark.scala#L153-L167),
with enabling and disabling this feature. Verified 20% wall clock time
improvement (switch control and test group order as well to verify the
improvement to not be the noise).
```
Running benchmark: shuffle hash join
Running case: shuffle hash join unique key SHJ off
Stopped after 5 iterations, 4039 ms
Running case: shuffle hash join unique key SHJ on
Stopped after 5 iterations, 2898 ms
Java HotSpot(TM) 64-Bit Server VM 1.8.0_181-b13 on Mac OS X 10.15.4
Intel(R) Core(TM) i9-9980HK CPU @ 2.40GHz
shuffle hash join: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
shuffle hash join unique key SHJ off 707 808
81 5.9 168.6 1.0X
shuffle hash join unique key SHJ on 547 580
50 7.7 130.4 1.3X
```
```
Running benchmark: shuffle hash join
Running case: shuffle hash join unique key SHJ on
Stopped after 5 iterations, 3333 ms
Running case: shuffle hash join unique key SHJ off
Stopped after 5 iterations, 4268 ms
Java HotSpot(TM) 64-Bit Server VM 1.8.0_181-b13 on Mac OS X 10.15.4
Intel(R) Core(TM) i9-9980HK CPU @ 2.40GHz
shuffle hash join: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
shuffle hash join unique key SHJ on 565 667
60 7.4 134.8 1.0X
shuffle hash join unique key SHJ off 774 854
85 5.4 184.4 0.7X
```
### Does this PR introduce _any_ user-facing change?
<!--
Note that it means *any* user-facing change including all aspects such as
the documentation fix.
If yes, please clarify the previous behavior and the change this PR proposes
- provide the console output, description and/or an example to show the
behavior difference if possible.
If possible, please also clarify if this is a user-facing change compared to
the released Spark versions or within the unreleased branches such as master.
If no, write 'No'.
-->
No.
### How was this patch tested?
<!--
If tests were added, say they were added here. Please make sure to add some
test cases that check the changes thoroughly including negative and positive
cases if possible.
If it was tested in a way different from regular unit tests, please clarify
how you tested step by step, ideally copy and paste-able, so that other
reviewers can test and check, and descendants can verify in the future.
If tests were not added, please describe why they were not added and/or why
it was difficult to add.
-->
* Added test in `OuterJoinSuite` to cover left outer and right outer join.
* Added test in `ExistenceJoinSuite` to cover left semi join.
* [Existing `joinSuite` already covered inner
join.](https://github.com/apache/spark/blob/master/sql/core/src/test/scala/org/apache/spark/sql/JoinSuite.scala#L182)
* [Existing `ExistenceJoinSuite` already covered left anti
join.](https://github.com/apache/spark/blob/master/sql/core/src/test/scala/org/apache/spark/sql/execution/joins/ExistenceJoinSuite.scala#L228)
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
