[
https://issues.apache.org/jira/browse/FLINK-13475?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16896226#comment-16896226
]
Stephan Ewen commented on FLINK-13475:
--------------------------------------
Hadoop dependencies are generally super messed up.
I would suggest to go through the Hive dependency and exclude whatever possible.
For example {{hadoop-common}} alone an insane transitive dependency tree with
webserver libraries, etc. that we surely don't want.
> Reduce dependency on third-party maven repositories
> ---------------------------------------------------
>
> Key: FLINK-13475
> URL: https://issues.apache.org/jira/browse/FLINK-13475
> Project: Flink
> Issue Type: Improvement
> Components: Connectors / Hive
> Affects Versions: 1.9.0
> Reporter: Till Rohrmann
> Priority: Critical
> Fix For: 1.9.0, 1.10.0
>
> Attachments: flink-connector-hive-dependency.txt
>
>
> A user reported that Flink's Hive connectors requires third-party maven
> repositories which are not everywhere accessible in order to build.
> Concretely, the hive connector requires access to Conjars for
> org.pentaho:pentaho-aggdesigner-algorithm and javax.jms:jms:jar:1.1.
> It would be great to reduce the dependency on third-party maven repositories
> if possible. For future reference, other projects faced similar problems:
> CALCITE-605, CALCITE-1474
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
