[
https://issues.apache.org/jira/browse/SPARK-2387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14270444#comment-14270444
]
Lianhui Wang edited comment on SPARK-2387 at 1/9/15 2:44 AM:
-------------------------------------------------------------
[~xuefuz] [~sandyr] [~lirui]
yes, i think firstly we provide a option to use this
feature.example:spark.scheduler.removeStageBarrier
if need this feature, we just open a option and can own performance
improvement. it is important that will not affect other's features.
was (Author: lianhuiwang):
[~xuefuz] [~sandyr] [~lirui]
yes, i think firstly we can provide a option to use this
feature.example:spark.scheduler.removeStageBarrier
if need this feature, we just open a option and can own performance
improvement. it is important that will not affect other's features.
> Remove the stage barrier for better resource utilization
> --------------------------------------------------------
>
> Key: SPARK-2387
> URL: https://issues.apache.org/jira/browse/SPARK-2387
> Project: Spark
> Issue Type: New Feature
> Components: Spark Core
> Reporter: Rui Li
>
> DAGScheduler divides a Spark job into multiple stages according to RDD
> dependencies. Whenever there’s a shuffle dependency, DAGScheduler creates a
> shuffle map stage on the map side, and another stage depending on that stage.
> Currently, the downstream stage cannot start until all its depended stages
> have finished. This barrier between stages leads to idle slots when waiting
> for the last few upstream tasks to finish and thus wasting cluster resources.
> Therefore we propose to remove the barrier and pre-start the reduce stage
> once there're free slots. This can achieve better resource utilization and
> improve the overall job performance, especially when there're lots of
> executors granted to the application.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
