Github user tgravescs commented on a diff in the pull request:
https://github.com/apache/spark/pull/3508#discussion_r21090918
--- Diff:
core/src/main/scala/org/apache/spark/scheduler/cluster/YarnSchedulerBackend.scala
---
@@ -92,6 +104,57 @@ private[spark] abstract class YarnSchedulerBackend(
}
/**
+ * This system security manager applies to the entire process.
+ * It's main purpose is to handle the case if the user code does a
System.exit.
+ * This allows us to catch that and properly set the YARN application
status and
+ * cleanup if needed.
+ */
+ private def setupSystemSecurityManager(amActor: ActorRef): Unit = {
--- End diff --
The securityManager in the AM was causing a performance impact and we just
removed it. I expect the same issue to happen here.
https://github.com/apache/spark/pull/3484
---
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]
