[
https://issues.apache.org/jira/browse/HIVE-15205?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15668286#comment-15668286
]
Vaibhav Gumashta commented on HIVE-15205:
-----------------------------------------
[~sershe] This jira doesn't touch ImportSemanticAnalyzer etc. It is resolving
some of the problems you are pointing to in original jira. All the classes here
are new (unless you have created ones with similar names).
> Create ReplDumpTask/ReplDumpWork for dumping out metadata
> ---------------------------------------------------------
>
> Key: HIVE-15205
> URL: https://issues.apache.org/jira/browse/HIVE-15205
> Project: Hive
> Issue Type: Sub-task
> Components: repl
> Reporter: Vaibhav Gumashta
> Assignee: Vaibhav Gumashta
>
> The current bootstrap code generates dump metadata during semantic analysis
> which breaks security and task/work abstraction. It also uses existing
> classes (from Export/Import world) for code reuse purpose, but as a result
> ends up dealing with a lot if-then-elses. It makes sense to have a cleaner
> abstraction which uses ReplDumpTask and ReplDumpWork (to configure the Task).
> Also perhaps worth evaluating ReplLoadTask/ReplLoadWork for load side.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
