[
https://issues.apache.org/jira/browse/MAHOUT-708?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13037699#comment-13037699
]
Dmitriy Lyubimov commented on MAHOUT-708:
-----------------------------------------
-1 in general.
Most folks are either on 0.20.2 (EMR) or CDH3 (baremetal). I know of no one
using 0.21. I am not sure that using 0.21 new api will be 100% compatible with
CDH3, there are still some missing pieces there. So if you move, you may have
me locked in 0.5 since i am a CDH3 user. (and EMR for bigger trains).
What i think might be reasonable is to create a branch with cdh3 dependencies
and make sure all tests are passing (i saw 2 or 3 not passing), albeit
generally everything compiles with cdh3. _Then we would cover all major camps
out there_ with practically same codebase.
Yes i am also waiting for new hadoop architecture to come out (i think they
were saying mid summer), a fundamental rewrite where task resource is separated
from a concept of application (i.e. map reduce) and that would really be great.
That would be a worthy update.
-d
> Update to Hadoop 0.21
> ---------------------
>
> Key: MAHOUT-708
> URL: https://issues.apache.org/jira/browse/MAHOUT-708
> Project: Mahout
> Issue Type: Task
> Components: Classification, Clustering, Collaborative Filtering,
> Frequent Itemset/Association Rule Mining
> Affects Versions: 0.5, 0.6
> Reporter: Sean Owen
> Assignee: Sean Owen
> Labels: hadoop
> Fix For: 1.0
>
>
> I suggest we should move to Hadoop 0.21 for the next release. It is the
> current release, soon to be superseded by 0.22. It matches more closely what
> CDH3/4 users use. It has bug fixes, and crucially some features that make
> joins much less painful.
> The drawback is that EMR does not yet support it. I still suggest we forge
> ahead as one imagines it will be supported by the time we release 0.6.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
