[
https://issues.apache.org/jira/browse/NUTCH-2049?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14701495#comment-14701495
]
Asitang Mishra commented on NUTCH-2049:
---------------------------------------
Hi Chris,
The Naive Bayes plugin, since has a hadoop job of it's own. does only work in
local mode and not distributed. Because, the Parse job of which this plugin is
a part, is also a hadoop job. So, it becomes a nested hadoop job.
Since, the training part of the plugin is the only one that is a hadoop job
(and not the classification). I can make a separate tool for training. And keep
only the classification part in the plugin, which is not a hadoop job (And have
tested this in distributed mode).
> Upgrade Trunk to Hadoop > 2.4 stable
> ------------------------------------
>
> Key: NUTCH-2049
> URL: https://issues.apache.org/jira/browse/NUTCH-2049
> Project: Nutch
> Issue Type: Improvement
> Components: build
> Reporter: Lewis John McGibbney
> Assignee: Lewis John McGibbney
> Labels: memex
> Fix For: 1.11
>
> Attachments: NUTCH-2049.patch, NUTCH-2049v2.patch
>
>
> Convo here - http://www.mail-archive.com/dev%40nutch.apache.org/msg18225.html
> I am +1 for taking trunk (or a branch of trunk) to explicit dependency on >
> Hadoop 2.6.
> We can run our tests, we can validate, we can fix.
> I will be doing validation on 2.X in paralegal as this is what I use on my
> own projects.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
