[
https://issues.apache.org/jira/browse/HCATALOG-587?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13555500#comment-13555500
]
Travis Crawford commented on HCATALOG-587:
------------------------------------------
Should have explicitly called this out in the previous comment. In
{{LauncherDelegator}} we see:
{code}
+ addDef(args, AppConfig.HADOOP_CHILD_JAVA_OPTS,
appConf.controllerMRChildOpts());
{code}
In general there are not getters for keys, which IMHO is a good thing? The
following would be more in line with what we do elsewhere:
{code}
addDef(args, AppConfig.HADOOP_CHILD_JAVA_OPTS,
appConf.get(AppConfig.TEMPLETON_CONTROLLER_MR_CHILD_OPTS));
{code}
> webhcat controller map task takes too much memory
> -------------------------------------------------
>
> Key: HCATALOG-587
> URL: https://issues.apache.org/jira/browse/HCATALOG-587
> Project: HCatalog
> Issue Type: Bug
> Components: webhcat-svr
> Reporter: Thejas M Nair
> Assignee: Thejas M Nair
> Fix For: 0.5, 0.6
>
> Attachments: HCATALOG-587.1.patch, HCATALOG-587.2.patch
>
>
> The webhcat map only task does not need lot of memory. But when a java
> application is started with -server option, it allocates the memory specified
> as max heap size (-Xmx in oracle jre) upfront.
> When combined with the memory used by pig or hive client, this can sometimes
> exceed the maximum memory limit for a map task (if configured).
> Being able to specify the java options for the map task will be useful to
> avoid this issue.
--
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
