[jira] [Commented] (SPARK-19146) Drop more elements when stageData.taskData.size > retainedTasks to reduce the number of times on call drop

2017-01-10 Thread Apache Spark (JIRA) 

[ 
https://issues.apache.org/jira/browse/SPARK-19146?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15814355#comment-15814355
 ] 

Apache Spark commented on SPARK-19146:
--

User 'wangyum' has created a pull request for this issue:
https://github.com/apache/spark/pull/16527

> Drop more elements when stageData.taskData.size > retainedTasks to reduce the 
> number of times on call drop
> --
>
> Key: SPARK-19146
> URL: https://issues.apache.org/jira/browse/SPARK-19146
> Project: Spark
>  Issue Type: Improvement
>  Components: Spark Core
>Affects Versions: 2.1.0
>Reporter: Yuming Wang
> Attachments: can-not-consume-taskEnd-events.jpg
>
>
> The performance of the 
> [{{drop}}|https://github.com/apache/spark/blob/v2.1.0/core/src/main/scala/org/apache/spark/ui/jobs/JobProgressListener.scala#L412]
>   function is bad.
> * Modify the code: 
> {code:java}
>   if (stageData.taskData.size > retainedTasks) {
> val start = System.currentTimeMillis()
> stageData.taskData = stageData.taskData.drop(stageData.taskData.size 
> - retainedTasks)
> logInfo(s"Time consuming: ${System.currentTimeMillis() - start}")
>   }
> {code}
> *  Time consuming
> {noformat}
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 156
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 145
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 148
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 159
> {noformat}
> My opinion is drop more elements when {{stageData.taskData.size > 
> retainedTasks}} to reduce the number of times on call {{drop}}.



--
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

[jira] [Commented] (SPARK-19146) Drop more elements when stageData.taskData.size > retainedTasks to reduce the number of times on call drop

2017-01-09 Thread Yuming Wang (JIRA) 

[ 
https://issues.apache.org/jira/browse/SPARK-19146?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15814207#comment-15814207
 ] 

Yuming Wang commented on SPARK-19146:
-

The activated tasks more and more and then {{dynamicAllocation}} will lose 
efficacy If {{SparkListenerTaskEnd}} events cannot be consumed. 

!can-not-consume-taskEnd-events.jpg!

> Drop more elements when stageData.taskData.size > retainedTasks to reduce the 
> number of times on call drop
> --
>
> Key: SPARK-19146
> URL: https://issues.apache.org/jira/browse/SPARK-19146
> Project: Spark
>  Issue Type: Improvement
>  Components: Spark Core
>Affects Versions: 2.1.0
>Reporter: Yuming Wang
> Attachments: can-not-consume-taskEnd-events.jpg
>
>
> The performance of the 
> [{{drop}}|https://github.com/apache/spark/blob/v2.1.0/core/src/main/scala/org/apache/spark/ui/jobs/JobProgressListener.scala#L412]
>   function is bad.
> * Modify the code: 
> {code:java}
>   if (stageData.taskData.size > retainedTasks) {
> val start = System.currentTimeMillis()
> stageData.taskData = stageData.taskData.drop(stageData.taskData.size 
> - retainedTasks)
> logInfo(s"Time consuming: ${System.currentTimeMillis() - start}")
>   }
> {code}
> *  Time consuming
> {noformat}
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 156
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 145
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 148
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 159
> {noformat}
> My opinion is drop more elements when {{stageData.taskData.size > 
> retainedTasks}} to reduce the number of times on call {{drop}}.



--
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

[jira] [Commented] (SPARK-19146) Drop more elements when stageData.taskData.size > retainedTasks to reduce the number of times on call drop

2017-01-09 Thread Yuming Wang (JIRA) 

[ 
https://issues.apache.org/jira/browse/SPARK-19146?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15814152#comment-15814152
 ] 

Yuming Wang commented on SPARK-19146:
-

I will create a PR later

> Drop more elements when stageData.taskData.size > retainedTasks to reduce the 
> number of times on call drop
> --
>
> Key: SPARK-19146
> URL: https://issues.apache.org/jira/browse/SPARK-19146
> Project: Spark
>  Issue Type: Improvement
>  Components: Spark Core
>Affects Versions: 2.1.0
>Reporter: Yuming Wang
>
> The performance of the 
> [{{drop}}|https://github.com/apache/spark/blob/v2.1.0/core/src/main/scala/org/apache/spark/ui/jobs/JobProgressListener.scala#L412]
>   function is bad.
> * Modify the code: 
> {code:java}
>   if (stageData.taskData.size > retainedTasks) {
> val start = System.currentTimeMillis()
> stageData.taskData = stageData.taskData.drop(stageData.taskData.size 
> - retainedTasks)
> logInfo(s"Time consuming: ${System.currentTimeMillis() - start}")
>   }
> {code}
> *  Time consuming
> {noformat}
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 156
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 145
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 148
> 17/01/10 14:04:05 INFO JobProgressListener: Time consuming: 159
> {noformat}
> My opinion is drop more elements when {{stageData.taskData.size > 
> retainedTasks}} to reduce the number of times on call {{drop}}.



--
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