[
https://issues.apache.org/jira/browse/DAFFODIL-1440?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16701171#comment-16701171
]
Michael Beckerle commented on DAFFODIL-1440:
--------------------------------------------
Badly needed for Spark users.
> JAPI/SAPI DataProcessor should be serializable
> ----------------------------------------------
>
> Key: DAFFODIL-1440
> URL: https://issues.apache.org/jira/browse/DAFFODIL-1440
> Project: Daffodil
> Issue Type: Bug
> Components: API
> Affects Versions: 2.0.0
> Reporter: Steve Lawrence
> Priority: Critical
>
> Some tools, such as Apache Spark, may need to serialize the DataProcessor to
> be able to send it to cluster nodes. Although we have the save/reload
> functions to support serialization, apache spark doesn't know how to use
> those, and just serializes the DataProcessor. So the SAPI/JAPI DataProcessors
> (which wrap the real DataProcessors) must also be made serializable.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
