hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sushanth Sowmyan (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HIVE-4524) Make the Hive HBaseStorageHandler work under HCat
Date Wed, 08 May 2013 20:19:15 GMT

     [ https://issues.apache.org/jira/browse/HIVE-4524?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Sushanth Sowmyan updated HIVE-4524:
-----------------------------------

    Attachment: hbh4.patch

Attaching patch that introduces WrapperStorageHandler in HCat which wraps native HiveStorageHandlers
and makes them work from within HCat.

Also, it includes one non-hcat fix to the HiveHBaseTableOutputFormat, where its super.getConf()
would throw an exception if called before checkOutputSpecs() was called - I had to refactor
out common code to ensure that didn't happen.
                
> Make the Hive HBaseStorageHandler work under HCat
> -------------------------------------------------
>
>                 Key: HIVE-4524
>                 URL: https://issues.apache.org/jira/browse/HIVE-4524
>             Project: Hive
>          Issue Type: Bug
>          Components: HBase Handler, HCatalog
>            Reporter: Sushanth Sowmyan
>            Assignee: Sushanth Sowmyan
>         Attachments: hbh4.patch
>
>
> Currently, HCatalog has its own HCatHBaseStorageHandler that extends from HBaseStorageHandler
to allow for StorageHandler support, and does some translations, like org.apache.mapred->org.apache.mapreduce
wrapping, etc. However, this compatibility layer is not complete in functionality as it still
assumes the underlying OutputFormat is a mapred.OutputFormat implementation as opposed to
a HiveOutputFormat implementation, and it makes assumptions about config property copies that
implementations of the HiveStorageHandler, such as the HBaseStorageHandler do not do.
> To fix this, we need to improve the ability for HCat to properly load native-hive-style
StorageHandlers.
> Also, since HCat has its own HBaseStorageHandler and we'd like to not maintain two separate
HBaseStorageHandlers, the idea is to deprecate HCat's storage handler over time, and make
sure that hive's HBaseStorageHandler works properly from HCat, and over time, have it reach
feature parity with the HCat one.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message