cloud-fan commented on a change in pull request #25620: [SPARK-25341][Core]
Support rolling back a shuffle map stage and re-generate the shuffle files
URL: https://github.com/apache/spark/pull/25620#discussion_r321777091
##########
File path:
core/src/main/scala/org/apache/spark/storage/ShuffleBlockFetcherIterator.scala
##########
@@ -570,11 +574,12 @@ final class ShuffleBlockFetcherIterator(
private[storage] def throwFetchFailedException(
blockId: BlockId,
+ mapIndex: Int,
address: BlockManagerId,
e: Throwable) = {
blockId match {
- case ShuffleBlockId(shufId, mapId, reduceId) =>
- throw new FetchFailedException(address, shufId.toInt, mapId.toInt,
reduceId, e)
+ case ShuffleBlockId(shufId, _, reduceId) =>
+ throw new FetchFailedException(address, shufId, mapIndex, reduceId, e)
Review comment:
shall we put `mapTaskId` in `FetchFailedException`? I take a look at the
code, and seems `FetchFailedException` is for logging purpose. I think it's
better to see `mapTaskId` and `reduceId`, so that we can know the shuffle block
id which can help debugging.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
