wankunde opened a new pull request, #36709:
URL: https://github.com/apache/spark/pull/36709
…mance
<!--
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?
Optimize `MapOutputTracker.convertMapStatuses()` method.
### Why are the changes needed?
`MapOutputTracker.convertMapStatuses()` will be very slow if there are tens
of thousands MapStatuses and MergeStatuses.
Benchmark code:
```java
val benchmark = new Benchmark("MapStatuses Convert", 1, output = output)
val blockManagerNumber = 1000
val mapNumber = 50000
val shufflePartitions = 10000
val shuffleId: Int = 0
// First reduce task will fetch map data from startPartition to
endPartition
val startPartition = 0
val startMapIndex = 0
val endMapIndex = mapNumber
val blockManagers = Array.tabulate(blockManagerNumber) { i =>
BlockManagerId("a", "host" + i, 7337)
}
val mapStatuses: Array[MapStatus] = Array.tabulate(mapNumber) {
mapTaskId =>
HighlyCompressedMapStatus(
blockManagers(mapTaskId % blockManagerNumber),
Array.tabulate(shufflePartitions)(i => if (i % 50 == 0) 1 else 0),
mapTaskId)
}
val bitmap = new RoaringBitmap()
Range(0, 4000).foreach(bitmap.add(_))
val mergeStatuses = Array.tabulate(shufflePartitions) { part =>
MergeStatus(blockManagers(part % blockManagerNumber), shuffleId,
bitmap, 100)
}
Array(499, 999, 1499).foreach { endPartition =>
benchmark.addCase(
s"Num Maps: $mapNumber Fetch partitions:${endPartition -
startPartition + 1}",
numIters) { _ =>
MapOutputTracker.convertMapStatuses(
shuffleId,
startPartition,
endPartition,
mapStatuses,
startMapIndex,
endMapIndex,
Some(mergeStatuses))
}
}
```
Before this PR
```
================================================================================================
MapStatuses Convert Benchmark
================================================================================================
Java HotSpot(TM) 64-Bit Server VM 1.8.0_281-b09 on Mac OS X 10.15.7
Intel(R) Core(TM) i9-9980HK CPU @ 2.40GHz
MapStatuses Convert: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
Num Maps: 50000 Fetch partitions:500 3393 3483
96 0.0 3393439257.0 1.0X
Num Maps: 50000 Fetch partitions:1000 6640 6772
121 0.0 6639654832.0 0.5X
Num Maps: 50000 Fetch partitions:1500 10035 10143
108 0.0 10035100069.0 0.3X
```
After this PR
```
================================================================================================
MapStatuses Convert Benchmark
================================================================================================
Java HotSpot(TM) 64-Bit Server VM 1.8.0_281-b09 on Mac OS X 10.15.7
Intel(R) Core(TM) i9-9980HK CPU @ 2.40GHz
MapStatuses Convert: Best Time(ms) Avg Time(ms)
Stdev(ms) Rate(M/s) Per Row(ns) Relative
------------------------------------------------------------------------------------------------------------------------
Num Maps: 50000 Fetch partitions:500 667 679
15 0.0 666562302.0 1.0X
Num Maps: 50000 Fetch partitions:1000 1285 1397
115 0.0 1284808865.0 0.5X
Num Maps: 50000 Fetch partitions:1500 2045 2068
32 0.0 2044951906.0 0.3X
```
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Exists UTs.
--
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]
