Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
MapReduce >> mail # user >> hadoop -Mapreduce


Copy link to this message
-
Re: hadoop -Mapreduce
Hi,

I am fresher to mapreduce concept,
I would like to know how to implement multithread concept in mapreduce.

Please provide sample program for implementating the above concept.

Thanks in advance.
Ranjini.R

On Mon, Dec 9, 2013 at 12:34 PM, Ranjini Rathinam <[EMAIL PROTECTED]>wrote:

> hi,
>
> Will these class support in hadoop 0.20 version. I am using this only.?
>
> Please suggest.
>
> Thanksin advance,
>
> Ranjini.R
>
>  On Mon, Dec 9, 2013 at 12:27 PM, Vinayakumar B <[EMAIL PROTECTED]>wrote:
>
>>  Hi Ranjini,
>>
>>
>>
>> Please check this mapper class whether this suites your needs.
>>
>> org.apache.hadoop.mapreduce.lib.map.MultithreadedMapper
>>
>>
>>
>> Thanks and Regards,
>>
>> Vinayakumar B
>>
>> *From:* Ranjini Rathinam [mailto:[EMAIL PROTECTED]]
>> *Sent:* 09 December 2013 11:13
>> *To:* [EMAIL PROTECTED]
>> *Subject:* Re: hadoop -Mapreduce
>>
>>
>>
>> HI,
>>
>>
>>
>> I need to use thread functionality to run more than one mapper class
>> parallelly.
>>
>>
>>
>> Please suggest with some sample code.
>>
>>
>>
>> Thanks in advance.
>>
>> On Fri, Dec 6, 2013 at 4:37 PM, Subroto <[EMAIL PROTECTED]> wrote:
>>
>> Hi Ranjini,
>>
>>
>>
>> The number of mappers depend on InputSplits and which intern depends on
>> size of input data.
>>
>> The number of reducers can be configured by "mapred.reduce.tasks"
>>
>>
>>
>> Further you can get more information on numbers of Maps and Reduce for a
>> Job from:
>>
>> http://wiki.apache.org/hadoop/HowManyMapsAndReduces
>>
>>
>>
>> Running parallel mappers and reducers depends on availability of map and
>> reduce slots in the cluster.
>>
>>
>>
>> Cheers,
>>
>> Subroto Sanyal
>>
>>
>>
>> On Dec 6, 2013, at 12:00 PM, Ranjini Rathinam wrote:
>>
>>
>>
>>  HI,
>>
>>
>>
>> How to run more than one mapper and reduce parallelly.?
>>
>>
>>
>> Please suggest.Thanks in advance.
>>
>>
>>
>> Ranjini.
>>
>>
>>
>>
>>
>
>