[
https://issues.apache.org/jira/browse/HADOOP-12615?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Wei-Chiu Chuang updated HADOOP-12615:
-------------------------------------
Attachment: HADOOP-12615.006.patch
Rev06:
Fixed the bug discovered by Zhe. Also, did a refactoring in MiniKMS which uses
a private helper method to error-handle copying resource input stream in a
consistent manner.
Similarly, created a convenient method getResourceAsStream() in MiniKdc to get
resource as stream (basically, replicate ThreadUtil.getResourceAsStream(),
because MiniKdc does not depend on Hadoop)
> Fix NPE in MiniKMS.start()
> --------------------------
>
> Key: HADOOP-12615
> URL: https://issues.apache.org/jira/browse/HADOOP-12615
> Project: Hadoop Common
> Issue Type: Bug
> Components: kms, test
> Affects Versions: 3.0.0
> Reporter: Wei-Chiu Chuang
> Assignee: Wei-Chiu Chuang
> Priority: Minor
> Labels: jenkins, supportability, test
> Attachments: HADOOP-12615.001.patch, HADOOP-12615.002.patch,
> HADOOP-12615.003.patch, HADOOP-12615.004.patch, HADOOP-12615.005.patch,
> HADOOP-12615.006.patch
>
>
> Sometimes, KMS resource file can not be loaded. When this happens, an
> InputStream variable will be a null pointer which will subsequently throw NPE.
> This is a supportability JIRA that makes the error message more explicit, and
> explain why NPE is thrown. Ultimately, leads us to understand why the
> resource files can not be loaded.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
