You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Sean R. Owen (Jira)" <ji...@apache.org> on 2019/10/26 22:49:00 UTC
[jira] [Resolved] (SPARK-29315) RDD.cache() called early creates
problems
[ https://issues.apache.org/jira/browse/SPARK-29315?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Sean R. Owen resolved SPARK-29315.
----------------------------------
Resolution: Invalid
You are modifying the cached objects in memory. This isn't going to work as you expect as those changes have undefined semantics.
> RDD.cache() called early creates problems
> -----------------------------------------
>
> Key: SPARK-29315
> URL: https://issues.apache.org/jira/browse/SPARK-29315
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Affects Versions: 2.4.4
> Environment: Apache Spark 2.4.4
> Windows 10
> Reporter: Edmond La Chance
> Priority: Minor
>
> First issue I post here. I noticed that when I call RDD.cache() early in my code, the results are all wrong!
> If I remove the call to cache(), or I add cache later in the code, after the first map transformation, it works fine.
> The graph is created from a data structure that already contains the random.
>
> I have posted versions that work, and versions that don't work here in this gist.
> [https://gist.github.com/mitchi/edd9637687cf47fac2616bb72932f8e7]
> here is an output that works :
> _Colors of the graph_
> _3 2 1 3 2 1 1 4 2 3_
> and an output that doesn't work :
> _Colors of the graph_
> _25 16 36 49 3 1 6 15 10 3_
>
>
--
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