[SPARK-8321][SQL][WIP] Authorization Support(on all operations not only DDL) in Spark Sql

[winningsix]

This PR enables Hive authorization V2(SQL based authorization). It adds authorization check after items are resolved. By converting logical plan node into HivePrivilegeObjects, we are able to leveraging Hive SQL based authorization.
The PR is marked WIP since it has limited supports on SQL operation and doesn't rebase onto the latest upstream. And used for reviews and sharing what in mind.

[AmplabJenkins]

Can one of the admins verify this patch?

[winningsix]

Hi @rxin, could you help review this PR? This PR is kind of POC code which is based on #10144 and serve as illustrating how to enable Hive authorization V2 in Spark SQL side. We can divide this task into several smaller sub tasks if we are in the right direction. Thank you!

[rxin]

Thanks for the pull request. I'm going through a list of pull requests to cut them down since the sheer number is breaking some of the tooling we have. Due to lack of activity on this pull request, I'm going to push a commit to close it. Feel free to reopen it or create a new one. We can also continue the discussion on the JIRA ticket.

[yaooqinn]

Hi @winningsix, I am interesting in your idea, and i am confusing about the field userName used in your code, 1)where is it be initialized? 2) is it used for privilege checking?

[winningsix]

@yaooqinn thank you for your interest. It's required by Hive side at SessionState. And it's not used for privilege check. Some companies have already used this pull request in their product environment with some modifications.

[yaooqinn]

@winningsix Glad to hear from you. And we can only use this through Spark ThriftServer, right?

[winningsix]

@yaooqinn yes, the validation is working on server side.

[zhaiyuyong]

how do this in spark with version 2.1.0》