[
https://issues.apache.org/jira/browse/HUDI-180?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16938144#comment-16938144
]
Nishith Agarwal edited comment on HUDI-180 at 9/26/19 12:10 AM:
----------------------------------------------------------------
As of now, we did not shed any dependencies but provided support for folks to
use metastore API in case a secure (kerberos) type of cluster setup is used.
In the long term, we need to make all the dependencies non-shaded so it can be
pulled in from runtime.
The ticket has been merged, closing it now.
was (Author: nishith29):
As of now, we did not shed any dependencies but provided support for folks to
use metastore API in case a secure (kerberos) type of cluster setup is used.
The ticket has been merged, closing it now.
> Move HiveSyncTool registration from hive server (jdbc) to metastore (thrift)
> ----------------------------------------------------------------------------
>
> Key: HUDI-180
> URL: https://issues.apache.org/jira/browse/HUDI-180
> Project: Apache Hudi (incubating)
> Issue Type: New Feature
> Components: Hive Integration
> Reporter: Nishith Agarwal
> Assignee: Nishith Agarwal
> Priority: Major
> Labels: pull-request-available
> Time Spent: 0.5h
> Remaining Estimate: 0h
>
> Currently, the HiveSyncTool takes in the JDBC URL of the hive server as an
> argument to be able to register hoodie datasets with the hive metastore.
> One of the problems faced was when using HiveSyncTool in a secure HDFS
> cluster environment using kerberos. The current implementation of JDBC does
> not allow for registration in such an environment. The implementation can be
> changed to support that but the consensus internally in our company has been
> to move to metastore.
> This ticket is to propose this change. Let's discuss on this ticket and I can
> follow this up with a PR.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
