Github user andrewor14 commented on a diff in the pull request:
https://github.com/apache/spark/pull/4887#discussion_r38582236
--- Diff: core/src/main/scala/org/apache/spark/CacheManager.scala ---
@@ -152,6 +152,9 @@ private[spark] class CacheManager(blockManager:
BlockManager) extends Logging {
*/
updatedBlocks ++=
blockManager.putIterator(key, values, level, tellMaster = true,
effectiveStorageLevel)
+ if (level.useMemory) {
+ blockManager.memoryStore.releasePendingUnrollMemoryForThisThread()
+ }
--- End diff --
@suyanNone this only happens if we failed to unroll in memory. In this case
we can't release the pending unroll memory because we created an `Array` that
contains some (but not all) of the values in this partitions. We need this
`Array` to return an iterator to the caller of this method so we can't release
it until after the task ends.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
