Deploying Mahout

["Ramani, Arun" <ar...@paypal.com>]

Hi,

I am planning to deploy Mahout on a large scale hadoop cluster. I have the following questions for which I was not able to find answers:

  *   Is Mahout installed only on the client machine (gateway to the hadoop cluster) ? Or do we need to push a jar on all the datanodes of a cluster
  *   Can Mahout be deployed on Hadoop-0.22 version?
  *   Can Mahout be deployed on Hadoop-1.0.3 version?
  *   If if it is merely a client side installation, how does it work on a distributed cluster.

It would be really helpful, If you could give me some insights on the above questions.

Thanks
Arun Ramani

Re: Deploying Mahout

[Sean Owen <sr...@gmail.com>]

MapReduce programs are never installed directly on a Hadoop cluster. Hadoop
deploys the program JAR to workers as needed. This is not specific to
Mahout.

Mahout compiles against 0.20.205 and so needs to be used with 0.20.205. It
will work with 1.0.3 as far as I know, with a recompile, as they are API
compatible.

I think 0.22 is not API compatible, no.



On Fri, Aug 17, 2012 at 1:30 AM, Ramani, Arun <ar...@paypal.com> wrote:

> Hi,
>
> I am planning to deploy Mahout on a large scale hadoop cluster. I have the
> following questions for which I was not able to find answers:
>
>   *   Is Mahout installed only on the client machine (gateway to the
> hadoop cluster) ? Or do we need to push a jar on all the datanodes of a
> cluster
>   *   Can Mahout be deployed on Hadoop-0.22 version?
>   *   Can Mahout be deployed on Hadoop-1.0.3 version?
>   *   If if it is merely a client side installation, how does it work on a
> distributed cluster.
>
> It would be really helpful, If you could give me some insights on the
> above questions.
>
> Thanks
> Arun Ramani
>