[batch-wordcount] Readme.md is unclear

[1123]

First of all thanks for these great examples. Very helpful.

The documentation of the batch-wordcount sample should be updated to

1. mention if the module can also be loaded via module upload --file ... --type job --name wordcount, instead of copying via script.

2. mention how to build against other versions of hadoop (e.g. 2.6). I suspect it would simply be sufficient to change the hadoop.version property in the pom.xml.

I tried to execute the wordcount sample as described above, but I get an exception when running the job: java.lang.ClassNotFoundException: org.apache.hadoop.examples.WordCount$TokenizerMapper