zuston commented on code in PR #358:
URL: https://github.com/apache/incubator-uniffle/pull/358#discussion_r1033085457
##########
client-spark/spark3/src/main/java/org/apache/spark/shuffle/reader/RssShuffleReader.java:
##########
@@ -206,7 +206,7 @@ public Configuration getHadoopConf() {
CreateShuffleReadClientRequest request = new
CreateShuffleReadClientRequest(
appId, shuffleId, partition, storageType, basePath,
indexReadLimit, readBufferSize,
1, partitionNum, partitionToExpectBlocks.get(partition),
taskIdBitmap, shuffleServerInfoList,
- hadoopConf, dataDistributionType);
+ hadoopConf, dataDistributionType, mapStartIndex, mapEndIndex);
Review Comment:
This is not precise.
I think I will introduce a new param to decide whether to use the
taskIdsBitmap, which will be specified by external invoker, like spark client.
And in spark client, we could use the endMapIndex to analyze the stage is data
skew.
--
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]
