c21 opened a new pull request #32242:
URL: https://github.com/apache/spark/pull/32242
<!--
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.
-->
For partial hash aggregation (code-gen path), we have two level of hash map
for aggregation. First level is from `RowBasedHashMapGenerator`, which is
computation faster compared to the second level from
`UnsafeFixedWidthAggregationMap`. The introducing of two level hash map can
help improve CPU performance of query as the first level hash map normally fits
in hardware cache and has cheaper hash function for key lookup.
For final hash aggregation, we can also support two level of hash map, to
improve query performance further.
The original two level of hash map code works for final aggregation mostly
out of box. The major change here is to support testing fall back of final
aggregation (see `isTestFinalAggregateWithFallback` and
`mergeFastHashMapForTest`).
Example:
An aggregation query:
```
spark.sql(
"""
|SELECT key, avg(value)
|FROM agg1
|GROUP BY key
""".stripMargin)
```
The generated code for final aggregation is
[here](https://gist.github.com/c21/d861bf7a1cece929a604f4bd77f3dc72).
An aggregation query with testing fallback:
```
withSQLConf("spark.sql.TungstenAggregate.testFallbackStartsAt" -> "2, 3") {
spark.sql(
"""
|SELECT key, avg(value)
|FROM agg1
|GROUP BY key
""".stripMargin)
}
```
The generated code for final aggregation is
[here](https://gist.github.com/c21/d0f704c0a33c24ec05387ff4df438bff). Note the
usage of `mergeFastHashMapForTest()`.
### 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.
-->
Improve the CPU performance of hash aggregation query in general.
For `AggregateBenchmark."Aggregate w multiple keys"`, seeing query
performance improved by 10%.
`codegen = T` means whole stage code-gen is enabled.
`hashmap = T` means two level maps is enabled for partial aggregation.
`finalhashmap = T` means two level maps is enabled for final aggregation.
```
Running benchmark: Aggregate w multiple keys
Running case: codegen = F
Stopped after 2 iterations, 8284 ms
Running case: codegen = T hashmap = F
Stopped after 2 iterations, 5424 ms
Running case: codegen = T hashmap = T finalhashmap = F
Stopped after 2 iterations, 4753 ms
Running case: codegen = T hashmap = T finalhashmap = T
Stopped after 2 iterations, 4508 ms
Java HotSpot(TM) 64-Bit Server VM 1.8.0_181-b13 on Mac OS X 10.15.7
Intel(R) Core(TM) i9-9980HK CPU @ 2.40GHz
Aggregate w multiple keys: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
codegen = F 3881 4142
370 5.4 185.1 1.0X
codegen = T hashmap = F 2701 2712
16 7.8 128.8 1.4X
codegen = T hashmap = T finalhashmap = F 2363 2377
19 8.9 112.7 1.6X
codegen = T hashmap = T finalhashmap = T 2252 2254
3 9.3 107.4 1.7X
```
### 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.
-->
Existing unit test in `HashAggregationQuerySuite` and
`HashAggregationQueryWithControlledFallbackSuite` already cover the test.
--
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]
