attilapiros commented on a change in pull request #25299: [SPARK-27651][Core] Avoid the network when shuffle blocks are fetched from the same host URL: https://github.com/apache/spark/pull/25299#discussion_r310178931
########## File path: core/src/main/scala/org/apache/spark/storage/BlockManagerMasterEndpoint.scala ########## @@ -51,6 +51,9 @@ class BlockManagerMasterEndpoint( // Mapping from block manager id to the block manager's information. private val blockManagerInfo = new mutable.HashMap[BlockManagerId, BlockManagerInfo] + // Mapping from executor id to the block manager's local disk directories. + private val executorIdToLocalDirs = new mutable.HashMap[String, Array[String]] Review comment: Yes, it grows without limit. Right now the code counts on there is a local dir for each of the requested executor so if we would like to limit this map the fetching logic must be changed too. But is not the same growing is present in the external shuffle service (even it could be worse there as it grows for each parallel Spark apps)? ---------------------------------------------------------------- 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]
