Github user squito commented on a diff in the pull request:
https://github.com/apache/spark/pull/21440#discussion_r191476566
--- Diff: core/src/main/scala/org/apache/spark/storage/BlockManager.scala
---
@@ -659,6 +659,11 @@ private[spark] class BlockManager(
* Get block from remote block managers as serialized bytes.
*/
def getRemoteBytes(blockId: BlockId): Option[ChunkedByteBuffer] = {
+ // TODO if we change this method to return the ManagedBuffer, then
getRemoteValues
+ // could just use the inputStream on the temp file, rather than
memory-mapping the file.
+ // Until then, replication can cause the process to use too much
memory and get killed
+ // by the OS / cluster manager (not a java OOM, since its a
memory-mapped file) even though
+ // we've read the data to disk.
--- End diff --
btw this fix is such low-hanging fruit that I would try to do this
immediately afterwards. (I haven't filed a jira yet just because there are
already so many defunct jira related to this, I was going to wait till my
changes got some traction).
I think its OK to get it in like this first, as this makes the behavior for
2.01 gb basically the same as it was for 1.99 gb.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
