This repository has been archived by the owner on Sep 18, 2023. It is now read-only.
Add a strategy to fall back to Vanilla Spark shuffle manager #1047
+898
−97
Conversation
This file contains 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
…ect#978) * [NSE-927] Add macro __AVX512BW__ check for different CPU architecture (oap-project#975) * Add __AVX512BW__ check * Fix cFormat * [NSE-126] set default codegen opt to O1 for branch-1.4
|
Thanks for opening a pull request! Could you open an issue for this pull request on Github Issues? https://github.com/oap-project/native-sql-engine/issues Then could you also rename commit message and pull request title in the following format? See also: |
|
Assuming the first two commits are not relevant to your patch, please do NOT include them. If your work depends on these commits, it would be better to open a dedicate PR to port them to main branch. |
PHILO-HE
changed the title
# Conflicts: # native-sql-engine/core/src/main/scala/com/intel/oap/expression/ConverterUtils.scala
zhouyuan
reviewed
Jul 28, 2022
| @@ -83,6 +83,11 @@ class GazellePluginConfig(conf: SQLConf) extends Logging { | |||
| val enableColumnarShuffledHashJoin: Boolean = | |||
| conf.getConfString("spark.oap.sql.columnar.shuffledhashjoin", "true").toBoolean && enableCpu | |||
|
|
|||
| // enable or disable fallback shuffle manager | |||
| val enableFallbackShuffle: Boolean = conf | |||
There was a problem hiding this comment.
can you please also add a short note on how to use this feature? and also make this default to false
There was a problem hiding this comment.
Added that in the description dialog
Disabled the configuration "spark.oap.sql.columnar.enableFallbackShuffle".
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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?
Add the strategy to fallback to Vanilla Spark shuffle manager.
o Enable fallback shuffle configuration and reuse the ColumnarShuffleExchangeExec
o Initiate the splitter iterator in Shuffle Dependency, and transform to the RDD: Produce2[Int, ColumnarBatch]
o Serialize the record batch to Shuffle Writer of Vanilla Spark.
How does this patch work?
When submit an application, we use native SQL engine with default ColumnarShuffleManager configuration,
--conf spark.shuffle.manager=org.apache.spark.shuffle.sort.ColumnarShuffleManagerHowever, we want to specify the custom or other shuffle manager for some situations, to enable Vanilla Spark shuffle manager，
--conf spark.shuffle.manager=org.apache.spark.shuffle.sort.SortShuffleManager --conf spark.oap.sql.columnar.enableFallbackShuffle=trueHow was this patch tested?
(Please explain how this patch was tested. E.g. unit tests, integration tests, manual tests)
(If this patch involves UI changes, please attach a screenshot; otherwise, remove this)