[jira] [Commented] (HIVE-13680) HiveServer2: Provide a way to compress ResultSets

["Alan Gates (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HIVE-13680?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15285371#comment-15285371 ] 

Alan Gates commented on HIVE-13680:
-----------------------------------

bq. If the server does not have that compressor plugin, it will request the jar from the client. The client should send the compressor plugin and all dependencies.
We should not do this.  It is a code injection security issue.  We cannot have users sending code into HS2 and the task nodes without full security controls.  In this case it makes more sense to make sure that we provide compressors for the standard options Hadoop already supports (snappy, lzo, ...)

Can the existing (de)compressors already in use in ORC etc. be used for this rather than re-implementing?

> HiveServer2: Provide a way to compress ResultSets
> -------------------------------------------------
>
>                 Key: HIVE-13680
>                 URL: https://issues.apache.org/jira/browse/HIVE-13680
>             Project: Hive
>          Issue Type: Sub-task
>          Components: HiveServer2, JDBC
>            Reporter: Vaibhav Gumashta
>            Assignee: Kevin Liew
>         Attachments: proposal.pdf
>
>
> With HIVE-12049 in, we can provide an option to compress ResultSets before writing to disk. The user can specify a compression library via a config param which can be used in the tasks.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)