yufeiyan1220 commented on a change in pull request #11953:
URL: https://github.com/apache/kafka/pull/11953#discussion_r837644305
##########
File path: core/src/main/scala/kafka/server/AbstractFetcherManager.scala
##########
@@ -62,19 +62,22 @@ abstract class AbstractFetcherManager[T <:
AbstractFetcherThread](val name: Stri
def resizeThreadPool(newSize: Int): Unit = {
def migratePartitions(newSize: Int): Unit = {
+ val allRemovedPartitionsMap = mutable.Map[TopicPartition,
InitialFetchState]()
fetcherThreadMap.forKeyValue { (id, thread) =>
- val partitionStates = removeFetcherForPartitions(thread.partitions)
+ val partitionStates = thread.removeAllPartitions()
+ failedPartitions.removeAll(partitionStates.keySet)
Review comment:
I am not very familiar with `FailedPartitions`, but I find that some
partitions are marked as `failedPartitions` and removed from fetchers. I think
the union of `FailedPartitions` and `removedFetcherForPartitions` is the whole
set of partitions. As the result, some partitions might not in
`failedPartitions`. I tend not to clear the `failedPartitions` there and just
remove all partitions that old thread possessed.
Another point is that in `addPartitions`, partitions in `FailedPartitions`
are removed before updating `partitionStates`. So maybe we can just remove this
line `failedPartitions.removeAll(partitionStates.keySet)`？
--
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]
