venkata91 commented on a change in pull request #33034:
URL: https://github.com/apache/spark/pull/33034#discussion_r676159400
##########
File path: core/src/main/scala/org/apache/spark/storage/BlockId.scala
##########
@@ -172,11 +195,15 @@ object BlockId {
val SHUFFLE_BATCH = "shuffle_([0-9]+)_([0-9]+)_([0-9]+)_([0-9]+)".r
val SHUFFLE_DATA = "shuffle_([0-9]+)_([0-9]+)_([0-9]+).data".r
val SHUFFLE_INDEX = "shuffle_([0-9]+)_([0-9]+)_([0-9]+).index".r
- val SHUFFLE_PUSH = "shufflePush_([0-9]+)_([0-9]+)_([0-9]+)".r
- val SHUFFLE_MERGED_DATA =
"shuffleMerged_([_A-Za-z0-9]*)_([0-9]+)_([0-9]+).data".r
- val SHUFFLE_MERGED_INDEX =
"shuffleMerged_([_A-Za-z0-9]*)_([0-9]+)_([0-9]+).index".r
- val SHUFFLE_MERGED_META =
"shuffleMerged_([_A-Za-z0-9]*)_([0-9]+)_([0-9]+).meta".r
- val SHUFFLE_CHUNK = "shuffleChunk_([0-9]+)_([0-9]+)_([0-9]+)".r
+ val SHUFFLE_PUSH = "shufflePush_([0-9]+)_(-?[0-9]+)_([0-9]+)_([0-9]+)".r
Review comment:
Yes we are not cleaning up the metadata from `RemoteBlockPushResolver`
for deterministic totally from the Map but will do it as a follow-up.
> Even if we were to do it, I would prefer to use 0 for deterministic stages
- and > 0 for others : including the very first stage attempt execution for a
non-deterministic stage.
Just curious, any reason specifically?
--
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]
