[
https://issues.apache.org/jira/browse/SPARK-19644?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15875702#comment-15875702
]
Deenbandhu Agarwal commented on SPARK-19644:
--------------------------------------------
I have analysed the issue more. I performed some of the experiments as follows
and analysed the heapdump using jvisualvm at some intervals.
1. Dstream.foreachRdd(rdd => rdd.map(x =>
someCaseClass(x)).take(10).foreach(println))
2. Dstream.foreachRdd(rdd => rdd.map(x => someCaseClass(x)).toDF.show(10,false))
3. Dstream.foreachRdd(rdd => rdd.map(x => someCaseClass(x)).toDS.show(10,false))
I Observed that the number of instances of
scala.collection.immutable.$colon$colon remain constant in 1 scenario but it
keeps on increasing in 2 and 3 scenario. So I think there is something leaky in
toDS or toDF function this may help you out to find out the issue.
> Memory leak in Spark Streaming
> ------------------------------
>
> Key: SPARK-19644
> URL: https://issues.apache.org/jira/browse/SPARK-19644
> Project: Spark
> Issue Type: Bug
> Components: DStreams
> Affects Versions: 2.0.2
> Environment: 3 AWS EC2 c3.xLarge
> Number of cores - 3
> Number of executors 3
> Memory to each executor 2GB
> Reporter: Deenbandhu Agarwal
> Priority: Critical
> Labels: memory_leak, performance
> Attachments: Dominator_tree.png, heapdump.png, Path2GCRoot.png
>
>
> I am using streaming on the production for some aggregation and fetching data
> from cassandra and saving data back to cassandra.
> I see a gradual increase in old generation heap capacity from 1161216 Bytes
> to 1397760 Bytes over a period of six hours.
> After 50 hours of processing instances of class
> scala.collection.immutable.$colon$colon incresed to 12,811,793 which is a
> huge number.
> I think this is a clear case of memory leak
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
