[jira] [Commented] (SPARK-979) Add some randomization to scheduler to better balance in-memory partition distributions

["Apache Spark (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-979?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14708739#comment-14708739 ] 

Apache Spark commented on SPARK-979:
------------------------------------

User 'tdas' has created a pull request for this issue:
https://github.com/apache/spark/pull/8387

> Add some randomization to scheduler to better balance in-memory partition distributions
> ---------------------------------------------------------------------------------------
>
>                 Key: SPARK-979
>                 URL: https://issues.apache.org/jira/browse/SPARK-979
>             Project: Spark
>          Issue Type: Improvement
>            Reporter: Reynold Xin
>            Assignee: Kay Ousterhout
>             Fix For: 1.0.0
>
>
> The Spark scheduler is very deterministic, which causes problems for the following workload (in serial order on a cluster with a small number of nodes):
> cache rdd 1 with 1 partition
> cache rdd 2 with 1 partition
> cache rdd 3 with 1 partition
> ....
> After a while, only executor 1 will have data in memory, and eventually leading to evicting in-memory blocks to disk while all other executors are empty. 
> We can solve this problem by adding some randomization to the cluster scheduling, or by adding memory aware scheduling (which is much harder to do). 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org