Hi,
I am trying to use MapReduce Pipes with version 0.20.1. of hadoop. If I
used org.apache.hadoop.mapreduce.lib.input.TextInputFormat as the input
format with the -InputFormat option, I am getting the following
error:
Exception in thread "main" java.lang.ClassCastException: class
org.apache.hadoop.mapreduce.lib.input.TextInputFormat
at java.lang.Class.asSubclass(Class.java:3018)
at
org.apache.hadoop.mapred.pipes.Submitter.getClass(Submitter.java:372)
at org.apache.hadoop.mapred.pipes.Submitter.run(Submitter.java:421)
at org.apache.hadoop.mapred.pipes.Submitter.main(Submitter.java:494)
I was able to run Pipes without the -InputFormat option, but my requirement
is to extend TextInputFormat. I guess Pipes is assuming the inputFormat to
be of type org.apache.hadoop.mapred.InputFormat, which is deprecated. So I
guess 0.20.1 is not fully ported to work with Pipes. Am I right here? Which
version should I use? Thanks.
Upendra
----- Original Message -----
From: "Erez Katz" <erez_k...@yahoo.com>
To: <common-dev@hadoop.apache.org>
Sent: Friday, November 27, 2009 6:59 PM
Subject: Re: mapreduce with non-text data
you could always base64 encode your binary data...
--- On Thu, 11/26/09, Upendra Dadi <ud...@gmu.edu> wrote:
From: Upendra Dadi <ud...@gmu.edu>
Subject: mapreduce with non-text data
To: common-dev@hadoop.apache.org
Date: Thursday, November 26, 2009, 5:07 PM
Hi,
Are there any use cases, examples of use of Hadoop
MapReduce for non-text data? Only examples that I see on the
web are for text data. Any pointers in that direction is
greatly appreciated. Thanks.
Regards,
Upendra
