[
https://issues.apache.org/jira/browse/HADOOP-12281?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Colin Patrick McCabe updated HADOOP-12281:
------------------------------------------
Comment: was deleted
(was: It seems like projects like Spark which use both Kryo and Hadoop will be
-1 on this change. Again, I feel like I'm missing something.
Isn't the classpath isolation work going to prevent problems like this? But
until it's complete, don't we just have to avoid doing things like this?)
> Add an incompatible version of Kryo to the hadoop-client classpath
> ------------------------------------------------------------------
>
> Key: HADOOP-12281
> URL: https://issues.apache.org/jira/browse/HADOOP-12281
> Project: Hadoop Common
> Issue Type: New Feature
> Components: build
> Reporter: Steve Loughran
> Assignee: Gunther Hagleitner
>
> Although versions of guava & jackson create compatibility problems, the issue
> of Kryo compatibility is put off until Hive, Spark and other libraries are
> loaded on the same classpath. By adding a version of Kryo to the Hadoop core
> libraries, we can break everything downstream in a consistent manner.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
