Input Output Parameters Error Hadoop

0
votes

I have a Mapper whose Input/Output parameters are:
(LongWritable, Text, WordPair, IntWritable)
and the reducer with Input/Output Parameters are:
(WordPair, IntWritable, WordPair, DoubleWritable)

I have set the JobConf as:
conf.setOutputKeyClass(WordPair.class);
conf.setOutputValueClass(DoubleWritable.class);
conf.setMapOutputValueClass(IntWritable.class);
conf.setOutputFormat(TextOutputFormat.class);

But I am getting error that:
-Java IOException DoubleWritable is not a class
-Map Job Failed

Also the job is failing before going to reducer phase. Is this part of my configuration correct? I believe I am doing some mismatch over here with the parameters.

Do I need to provide more Information? I am doing relative frequency problem in which I need to output from reducer values of type Double & I am using Old Api's.

1
hadoopmapreduce
can you provide the complete stacktrace of the exception? – Amar
Amar thanks for your reply ..Just now I figure out that this problem is coming because in the configuration I have mentioned my combiner class same as that of reducer class so that part is throwing exception as combiner input/output need to be same as map input/output. Now how to close this question as this is resolved? – – JackSparrow
leave it... though not much likely but it is possible that someone else might be doing the same mistake in future... – Amar

1 Answers

0
votes

I have figure out this question, I know I am replying to my own question but this is only because I want to let others know if they have the same problem.

This problem is coming because I have mentioned my combiner class as same as that of reducer class but as per this problem Input/output of mapper is not same as Input/Output of reducer whereas combiner should have same Input/output as that of mapper.

So I removed the line where I have set the combiner and It worked fine.