[
https://issues.apache.org/jira/browse/HBASE-6969?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13472973#comment-13472973
]
stack commented on HBASE-6969:
------------------------------
As written, at end of tests, won't your passed in zkcluster be shutdown? That
is probably not what you want? Should there be a getMiniZookeeperCluster to
answer your added setMiniZookeeperCluster? Is baseZKCluster necessary? Why
not just an internal flag which has whether or not HBaseTestingUtility started
the zk cluster? If we didn't start it, we shouldn't stop it on the way out?
Otherwise, looks like useful functionality to add. Thanks Micah.
> Allow for injecting MiniZookeeperCluster instance
> -------------------------------------------------
>
> Key: HBASE-6969
> URL: https://issues.apache.org/jira/browse/HBASE-6969
> Project: HBase
> Issue Type: Bug
> Components: test
> Affects Versions: 0.92.1
> Reporter: Micah Whitacre
> Attachments: hbase-6969.patch
>
>
> While not an official part of the HBase API, it'd be nice if
> HBaseTestingUtility allowed for consumer to inject their own instance of the
> MiniZookeeperCluster when executing tests.
> Currently there is no way to control how the ZK instance is started which
> we've had to hack around when there were ZK version conflicts (3.3 vs 3.4).
> Or if you want to control which specific port it starts on vs random.
> Allowing consumers to inject an instance (unstarted) gives them the freedom
> to implement their own without having to fork the entire HBaseTestingUtility
> class
--
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
