You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@paimon.apache.org by "wxplovecc (via GitHub)" <gi...@apache.org> on 2023/03/30 07:21:03 UTC
[GitHub] [incubator-paimon] wxplovecc opened a new pull request, #770: [flink] Assign batch splits for ContinuousFileSplitEnumerator
wxplovecc opened a new pull request, #770:
URL: https://github.com/apache/incubator-paimon/pull/770
*(Please specify the module before the PR name: [core] ... or [flink] ...)*
### Purpose
This PR solves #722 .
introduces an option scan.split-enumerator.batch-size for assigning splits at one time to avoid exceed akka.framesize limit in StaticFileStoreSplitEnumerator. ContinuousFileSplitEnumerator should support this mechanism too.
### Tests
`ContinuousFileSplitEnumeratorTest.testSplitWithBatch`
--
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: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-paimon] JingsongLi merged pull request #770: [flink] Assign batch splits for ContinuousFileSplitEnumerator
Posted by "JingsongLi (via GitHub)" <gi...@apache.org>.
JingsongLi merged PR #770:
URL: https://github.com/apache/incubator-paimon/pull/770
--
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: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-paimon] JingsongLi commented on a diff in pull request #770: [flink] Assign batch splits for ContinuousFileSplitEnumerator
Posted by "JingsongLi (via GitHub)" <gi...@apache.org>.
JingsongLi commented on code in PR #770:
URL: https://github.com/apache/incubator-paimon/pull/770#discussion_r1152896661
##########
paimon-flink/paimon-flink-common/src/main/java/org/apache/paimon/flink/source/ContinuousFileSplitEnumerator.java:
##########
@@ -205,9 +210,13 @@ private Map<Integer, List<FileStoreSourceSplit>> createAssignment() {
readersAwaitingSplit.remove(task);
return;
}
+ List<FileStoreSourceSplit> taskAssignment = new ArrayList<>();
+ while (!splits.isEmpty() && taskAssignment.size() < splitBatchSize) {
Review Comment:
I think it is OK to just assign one split to a task for a bucket.
Maybe here we can just limit assignment batch.
```
List<FileStoreSourceSplit> taskAssignment = assignment.computeIfAbsent(task, i -> new ArrayList<>());
if (taskAssignment.size() < splitBatchSize) {
taskAssignment.add(splits.poll());
}
```
This can be limited to a large number of buckets, and only a few tasks can work properly.
For example, buckets = 1000, but there is only one task.
--
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: issues-unsubscribe@paimon.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org