UnsupportedClassVersionError while running Mapreduce program

Question

When I am trying to run Mapreduce program, I am getting UnsupportedClassVersionError error.

Also when I am adding the following line for JOB 

Job job = new Job.getInstance(conf,"WordCount"); 

I am getting the error in my eclipse. To remove the error, I replace the above line with 

Job job = new Job(conf,"WordCount"); 

But for both the cases when I am trying to run the jar I am getting the error mentioned. Below is the Screenshot

Answer 1

Seems like the jar file was not exported properly. Did you export the jar by yourself or copied from somewhere? This is one of the many reasons for this error. Anyways, you can follow the below mentioned steps to run the MapReduce program:

1) Write the program in eclipse,

2) Create a jar file of this program,

Now, you will have the jar file in the local file system of your System.

3) Make sure that you have the input dataset in the hdfs,

4) Execute the Hadoop jar command.

The syntax to execute the hadoop jar command is as below,

hadoop jar <jarfilename.jar> <packagename.classname> <inputfile> <outputpath>

We will execute the below command for our program,

hadoop jar WordCount.jar co.edureka.WordCount wordcountproblem WordCountOutput

This will create a directory in hdfs named WordCountOutput and this directory will contain the output of our program,

5) Check the output.

The syntax to check the contents of a file in hdfs is as below,

hdfs dfs -cat <path/filename>

We will check the output of our program with the below command,

hdfs dfs -cat WordCountOutput/part-r-00000

You can see that it works without error.

Comments

Yes. It works. The first time, I used the jar file that I downloaded from the internet.