wankunde opened a new pull request, #38887:
URL: https://github.com/apache/spark/pull/38887
<!--
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'.
8. If you want to add or modify an error type or message, please read the
guideline first in
'core/src/main/resources/error/README.md'.
-->
### What changes were proposed in this pull request?
We can reorder window partition expressions by the distinct values stats.
Sorting with high cardinality will be faster.
Benchmark
```
object WindowBenchmark extends SqlBasedBenchmark {
import spark.implicits._
override def runBenchmarkSuite(mainArgs: Array[String]): Unit = {
val rand = new Random()
spark.range(1, 10000000, 1, 10)
.map(i => (i, i % 100, s"name-${rand.nextInt(100000)}")).toDF("a",
"b", "c")
.repartition(10, 'c)
.write.format("parquet").saveAsTable("t")
spark.sql(s"ANALYZE TABLE t COMPUTE STATISTICS FOR COLUMNS a, b")
runBenchmark("Window benchmark") {
val benchmark = new Benchmark("Window", 10000, 3, output = output)
val query = "SELECT *, row_number() over(partition by b, a order by c)
as rn FROM t"
benchmark.addCase("Without reorder window partition expressions") { _
=>
spark.sql(query).noop()
}
benchmark.addCase("With reorder window partition expressions") { _ =>
withSQLConf(
SQLConf.CBO_ENABLED.key -> "true",
SQLConf.REORDER_WINDOW_PARTITION_EXPRESSIONS.key -> "true") {
spark.sql(query).noop()
}
}
benchmark.run()
}
}
}
```
Benchmark result:
```
Window: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
----------------------------------------------------------------------------------------------------------------------------
Disable reorder window partition expressions 13338 14035
1138 0.0 1333756.2 1.0X
Enable reorder window partition expressions 5981 6030
77 0.0 598102.1 2.2X
```
### Why are the changes needed?
Improve Window operation
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Added UT
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
