質問

Does "Wordcount" program differs for different Hadoop(Mapreduce) versions. atleast for MRv1(mapreduce) and MRv2(YARN)?
or
Programming is same but needed packages for wordcount is varying.......?

役に立ちましたか?

解決

If you are talking about the WordCount example on Hadoop wiki, then there is no difference. You could either compile and run with MRv1 and MRv2 without modify any single line of the source code. All you need is to include correct class path as below:

javac -cp `hadoop classpath` WordCount.java

One thing to keep in mind is WordCount is just a HelloWorld example for Hadoop, which means you could implement it in various ways as you want. In fact, you could see the implementations in Hadoop 2.1.0 is already different from what you see on the Hadoop wiki

ライセンス: CC-BY-SA帰属
所属していません StackOverflow
scroll top