How does the 'hadoop' command work?

[Prateek Jindal <ji...@illinois.edu>]

Hello Hadoop users,

I am new to MapReduce. I had the following question. Please help me
regarding this.

"I am trying to run a very basic mapReduce application.

I copied *.java files (Mapper, reducer and the code to run the application)
into my home directory on hadoop cluster. Then when I run the following
command:

hadoop MaxTemperature input/ncdc/sample.txt output (here MaxTemperature is
the java file which calls JobClient).

I get the error that MaxTemperature.class is not found."

Could you suggest something about that? Does the "hadoop" command take the
".class file" or the ".java file"?

What command is used to start the map-reduce job?

Thanks,
Prateek.

Re: How does the 'hadoop' command work?

[abhishek sharma <ab...@usc.edu>]

Did you add the path to your home directory where the
MaxTemperature.class is into the Hadoop classpath?
I guess hadoop-env.sh  in ${HADOOP_HOME}/conf would be one place to do it.

Abhishek

On Fri, Feb 5, 2010 at 1:14 PM, Prateek Jindal <ji...@illinois.edu> wrote:
> Hello Hadoop users,
>
> I am new to MapReduce. I had the following question. Please help me
> regarding this.
>
> "I am trying to run a very basic mapReduce application.
>
> I copied *.java files (Mapper, reducer and the code to run the application)
> into my home directory on hadoop cluster. Then when I run the following
> command:
>
> hadoop MaxTemperature input/ncdc/sample.txt output (here MaxTemperature is
> the java file which calls JobClient).
>
> I get the error that MaxTemperature.class is not found."
>
> Could you suggest something about that? Does the "hadoop" command take the
> ".class file" or the ".java file"?
>
> What command is used to start the map-reduce job?
>
> Thanks,
> Prateek.
>