yuchenhuo commented on a change in pull request #34265:
URL: https://github.com/apache/spark/pull/34265#discussion_r728805893
##########
File path: core/src/main/scala/org/apache/spark/scheduler/DAGScheduler.scala
##########
@@ -841,6 +870,11 @@ private[spark] class DAGScheduler(
"Total number of partitions: " + maxPartitions)
}
+ // SPARK-23626: `RDD.getPartitions()` can be slow, so we eagerly compute
+ // `.partitions` on every RDD in the DAG to ensure that `getPartitions()`
+ // is evaluated outside of the DAGScheduler's single-threaded event loop:
+ eagerlyComputePartitionsForRddAndAncestors(rdd)
Review comment:
Would it be a good idea to add an assertion in the `DebugFilesystem` we
have to check that it's not accessed within the event loop thread? It might
help catch other cases where event loop might be doing heavy blocking operation.
--
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]
