[
https://issues.apache.org/jira/browse/HBASE-8124?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13603811#comment-13603811
]
Ted Yu commented on HBASE-8124:
-------------------------------
Yeah:
{code}
$ find . -name '*.java' -exec grep '"tmpjars' {} \; -print
conf.set("tmpjars",
String jars = conf.get("tmpjars");
./src/core/org/apache/hadoop/util/GenericOptionsParser.java
String libjars = job.get("tmpjars");
./src/mapred/org/apache/hadoop/mapred/JobClient.java
{code}
> TableMapReduceUtils should distribute dependency jars using an hbase config
> ---------------------------------------------------------------------------
>
> Key: HBASE-8124
> URL: https://issues.apache.org/jira/browse/HBASE-8124
> Project: HBase
> Issue Type: Sub-task
> Components: mapreduce, Usability
> Reporter: Nick Dimiduk
> Assignee: Nick Dimiduk
> Fix For: 0.95.0, 0.98.0
>
>
> When users invoke TableMapReduceUtils#addDependencyJars, directly or
> indirectly, the discovered jars are added to the jobconf via the
> configuration point "tmpjars". That means that, if I want to use one jobconf
> as the basis of another, and I blanket copy all "hbase.*" confs, the
> dependency jar list will be missed. Since the dependency jars are critical
> for successful execution of the job, and not an internal API passed between
> different MR components, this config point should be renamed to something
> starting with "hbase."
--
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
