[
https://issues.apache.org/jira/browse/HIVE-15446?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Chaoyu Tang updated HIVE-15446:
-------------------------------
Status: Patch Available (was: Open)
The old HADOOP_CLIENT_OPTS with main debug options should be removed from
HADOOP_OPTS when a new HADOOP_CLIENT_OPTS with child debug options will be
appended to HADOOP_OPTS when child process is invoked by HADOOP.sh.
[~ashutoshc] could review the patch? thanks
> Hive fails in recursive debug
> -----------------------------
>
> Key: HIVE-15446
> URL: https://issues.apache.org/jira/browse/HIVE-15446
> Project: Hive
> Issue Type: Bug
> Components: Diagnosability
> Reporter: Chaoyu Tang
> Assignee: Chaoyu Tang
> Priority: Minor
> Attachments: HIVE-15446.patch
>
>
> When running hive recursive debug mode, for example,
> ./bin/hive --debug:port=10008,childSuspend=y
> It fails with error msg:
> --
> ERROR: Cannot load this JVM TI agent twice, check your java command line for
> duplicate jdwp options.Error occurred during initialization of VM
> agent library failed to init: jdwp
> --
> It is because HADOOP_OPTS and HADOOP_CLIENT_OPTS both have jvm debug options
> when invoking HADOOP.sh for the child process. The HADOOP_CLIENT_OPTS is
> appended to HADOOP_OPTS in HADOOP.sh which leads to the duplicated debug
> options.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
