Sending the entire file content as value to the mapper

[Kasi Subrahmanyam]

Hi Team,

I have a file which has semi structured text  data with no definite start
and end points.
How can i send the entire content of the file at once as key or value to
the mapper instead of line by line.

Thanks,
Subbu

RE: Sending the entire file content as value to the mapper

[Devaraj k]

Hi,

  You could send the file meta info to the map function as key/value through 
the split, and then you can read the entire file in your map function.

Thanks
Devaraj k


-Original Message-
From: Kasi Subrahmanyam [mailto:kasisubbu...@gmail.com] 
Sent: 11 July 2013 13:38
To: common-user@hadoop.apache.org; mapreduce-u...@hadoop.apache.org
Subject: Sending the entire file content as value to the mapper

Hi Team,

I have a file which has semi structured text  data with no definite start and 
end points.
How can i send the entire content of the file at once as key or value to the 
mapper instead of line by line.

Thanks,
Subbu

Re: Task failure in slave node

[devara...@huawei.com]

Hi,

   It seems mahout-examples-0.7-job.jar is depending on other jars/classes.
While running Job Tasks it is not able to find those classes in the
classpath and failing those tasks.

You need to provide the dependent jar files while submitting/running Job.


Thanks
Devaraj k




--
View this message in context: 
http://lucene.472066.n3.nabble.com/Task-failure-in-slave-node-tp4077284p4077290.html
Sent from the Hadoop lucene-users mailing list archive at Nabble.com.