hadoop-mapreduce-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From swathi v <swatz...@gmail.com>
Subject Re: Streaming in mapreduce
Date Sat, 16 Jun 2012 10:37:13 GMT
Hi Pedro,

Adding to the response of *Bejoy*,

   - Hadoop streaming provides the user with the ability to use arbitrary
   programs in other languages like ruby, python, for a job’s map and reduce
   methods.
   - Streaming provides the ability to use external programs as any of the
   job’s mapper, combiner, or reducer.
   - The job is a traditional MapReduce job, with the framework handling
   input splitting, scheduling map tasks, scheduling input split pairs to run,
   shuffling and sorting map outputs, scheduling reduce tasks to run, and then
   writing the reduce output to the Hadoop Distributed File System (HDFS).
   - The framework handles a streaming job like any other MapReduce job.
   - The job might specify that an executable be used as the map processor
   and for the reduce processor.
   - Each task will start an instance of the applicable executable and
   write an applicable representation of the input key/value pairs to the
   executable.
   - The standard output of the executable is parsed as textual key/value
   pairs.
   - The executable being run for the reduce task will given an input line
   for each value in the reduce value iterator, composed of the key and
   that value. This link explains the same:
   http://wiki.apache.org/hadoop/HadoopStreaming (same as the response
   given by Ruslan Al-Fakikh)
   - *EXAMPLE:*The Hadoop Core distribution provides a Jython example
   MapReduce application in *src/examples/python/WordCount.py*
   - FYI : There are libraries available for C++. The C++ interface lends
   itself to usage by Simplified Wrapper and Interface Generator (SWIG) to
   generate other language interfaces. The usage of* Hadoop Pipes* and its
   example goes here: http://wiki.apache.org/hadoop/C++WordCount

Hope you find this useful.  :)
Thank You.

On Sat, Jun 16, 2012 at 3:00 PM, Bejoy KS <bejoy.hadoop@gmail.com> wrote:

> Hi Pedro
>
> In simple terms Streaming API is used in hadoop if you have your mapper or
> reducer is in any language other than java . Say ruby or python.
>
> Regards
> Bejoy KS
>
> Sent from handheld, please excuse typos.
> ------------------------------
> *From: * Pedro Costa <psdc1978@gmail.com>
> *Date: *Sat, 16 Jun 2012 10:23:20 +0100
> *To: *mapreduce-user@hadoop.apache.org<mapreduce-user@hadoop.apache.org>
> *ReplyTo: * mapreduce-user@hadoop.apache.org
> *Subject: *Re: Streaming in mapreduce
>
> I still don't get why hadoop streaming is useful. If I have man and reduce
> functions defined in shell script, like the one below, why should I use
> Hadoop?
>
> cat someInputFile | shellMapper.sh | shellReducer.sh > someOutputFile
>
>
>
> On 16/06/2012, at 01:21, Ruslan Al-Fakikh <metaruslan@gmail.com> wrote:
>
> Hi Pedro,
>
> You can find it here
> http://wiki.apache.org/hadoop/HadoopStreaming
>
> Thanks
>
> On Sat, Jun 16, 2012 at 2:46 AM, Pedro Costa <psdc1978@gmail.com> wrote:
>
> Hi,
>
>
> Hadoop mapreduce can be used for streaming. But what is streaming from the
> point of view of mapreduce? For me, streaming are video and audio data.
>
>
>  Why mapreduce supports streaming?
>
>
> Can anyone give me an example on why to use streaming in mapreduce?
>
>
> Thanks,
>
> Pedro
>
>


-- 
- Regards,
Swathi.V. ,
Software Developer
Blog URL :http://femgeekz.blogspot.in

Mime
View raw message