[
https://issues.apache.org/jira/browse/CHUKWA-473?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12855458#action_12855458
]
Bill Graham commented on CHUKWA-473:
------------------------------------
I've got a patch ready that makes the default map processor configurable, but
just wanted to clarify how we want to do the same for the reducer processor. It
looks for a class based on key.getReduceType() and falls back on
IdentityReducer in ReduceProcessorFactory. Assuming that's what we want to be
able to set a default for.
If so I need to change the signature of
{{ReduceProcessorFactory.getProcessor(String reduceType)}} to either take a
JobConf object or a default processor string, since I don't think that method
should return null and delegate the default logic to the caller. I'm leaning
slightly towards the later. Any opinions either way?
> Make default processor configurable
> -----------------------------------
>
> Key: CHUKWA-473
> URL: https://issues.apache.org/jira/browse/CHUKWA-473
> Project: Hadoop Chukwa
> Issue Type: New Feature
> Reporter: Bill Graham
> Assignee: Bill Graham
>
> The default processor is hard coded to {{DefaultProcessor}} in the {{Demux}}
> class. This should be configurable.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
