You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Hyukjin Kwon (Jira)" <ji...@apache.org> on 2021/05/03 10:44:00 UTC

[jira] [Commented] (SPARK-35251) Improve LiveEntityHelpers.newAccumulatorInfos performace

    [ https://issues.apache.org/jira/browse/SPARK-35251?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17338303#comment-17338303 ] 

Hyukjin Kwon commented on SPARK-35251:
--------------------------------------

Can you show reproducer and how you managed to get the profile?

> Improve LiveEntityHelpers.newAccumulatorInfos performace
> --------------------------------------------------------
>
>                 Key: SPARK-35251
>                 URL: https://issues.apache.org/jira/browse/SPARK-35251
>             Project: Spark
>          Issue Type: Improvement
>          Components: Spark Core
>    Affects Versions: 3.2.0
>            Reporter: Yuming Wang
>            Priority: Major
>
> [It|https://github.com/apache/spark/blob/3854ad87c78f2a331f9c9c1a34f9ec281900f8fe/core/src/main/scala/org/apache/spark/status/LiveEntity.scala#L646-L661] will impact performance if  there are lot of {{AccumulableInfo}} instances.
> {noformat}
>  num     #instances         #bytes  class name
> ----------------------------------------------
>    1:     331897850    50708371592  [C
>    2:        795358    24591381896  [J
>    3:     330836234    10586759488  java.lang.String
>    4:     139726297     8942483008  org.apache.spark.sql.execution.metric.SQLMetric
>    5:     249040156     5976963744  scala.Some
>    6:     146278719     5851148760  org.apache.spark.util.AccumulatorMetadata
>    7:      38448113     5536528272  java.net.URI
>    8:      37540162     3603855552  org.apache.hadoop.fs.FileStatus
>    9:      69724130     3346758240  java.util.Hashtable$Entry
>   10:      61521559     2953034832  java.util.concurrent.ConcurrentHashMap$Node
>   11:      50421974     2823630544  scala.collection.mutable.LinkedEntry
>   12:      43349222     2774350208  org.apache.spark.scheduler.AccumulableInfo
> {noformat}
> {noformat}
> --- 15430388364 ns (2.03%), 1543 samples
>   [ 0] scala.collection.TraversableLike.noneIn$1
>   [ 1] scala.collection.TraversableLike.filterImpl
>   [ 2] scala.collection.TraversableLike.filterImpl$
>   [ 3] scala.collection.AbstractTraversable.filterImpl
>   [ 4] scala.collection.TraversableLike.filter
>   [ 5] scala.collection.TraversableLike.filter$
>   [ 6] scala.collection.AbstractTraversable.filter
>   [ 7] org.apache.spark.status.LiveEntityHelpers$.newAccumulatorInfos
>   [ 8] org.apache.spark.status.LiveTask.doUpdate
>   [ 9] org.apache.spark.status.LiveEntity.write
> {noformat}



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

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