Mappers output collector

[maha <ma...@umail.ucsb.edu>]

This is confusing me ...

  Reported Maps # of output records is 0 , yet I see 86 bytes produced from each mapper in output destination path. 

-rw-r--r--   1 Hadoop supergroup         86 2011-03-22 10:17 /tmp/out/part-00000
.....
.....

 I can't eliminate output path because I get:   java.io.IOException: Undefined job output-path

How can I tell the job configuration not to prepare an output path (or anything produced by output.collect()) ?

Thank you,
Maha

Re: Mappers and RecordReaders

[maha <ma...@umail.ucsb.edu>]

Thanks Harsh :) I'll try it. Also, I'm wondering which of the following is true:

When I set a whole file as a split for a mapper, 

1) is the file only opened and the record reader reads from disk record by record ?

2) Or Part of the file is read into memory and the record reader get the record-by-record from memory ?

Assuming the single-thread mapper class.

Thanks,
Maha


On Mar 22, 2011, at 11:22 AM, Harsh J wrote:

> NullOutputFormat

Re: Mappers output collector

[Harsh J <qw...@gmail.com>]

Use the NullOutputFormat class as your OutputFormat, perhaps?

On Tue, Mar 22, 2011 at 11:21 PM, maha <ma...@umail.ucsb.edu> wrote:
> This is confusing me ...
>
>  Reported Maps # of output records is 0 , yet I see 86 bytes produced from each mapper in output destination path.
>
> -rw-r--r--   1 Hadoop supergroup         86 2011-03-22 10:17 /tmp/out/part-00000
> .....
> .....
>
>  I can't eliminate output path because I get:   java.io.IOException: Undefined job output-path
>
> How can I tell the job configuration not to prepare an output path (or anything produced by output.collect()) ?
>
> Thank you,
> Maha



-- 
Harsh J
http://harshj.com