hadoop-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Grant Mackey <gmac...@cs.ucf.edu>
Subject Re: Chaining Map-Reduce
Date Tue, 06 Jul 2010 19:15:43 GMT
Oh, of course it will. From what I know about using this it just  
allows you to launch 1 job instead of several. Every time a Map/Reduce  
pair finishes it will dump to the HDFS (or whatever you're using)

  - Grant

Quoting abc xyz <fabc_xyz111@yahoo.com>:

> one option can be to use multiple chained jobs using JobControl object.
> But can anyone tell whether in this case the output of first job be  
> written to
> the disk? Actually I want to minimize access to the disk.
>
> Cheers
>
>
>
> ________________________________
> From: Fariha Atta <fariha.atta.pk@gmail.com>
> To: general@hadoop.apache.org
> Sent: Fri, July 2, 2010 8:45:12 AM
> Subject: Re: Chaining Map-Reduce
>
> But it provides chaining like Map+ | Reduce | Map* i.e. we can use only one
> reducer.
>
>
> On Fri, Jul 2, 2010 at 8:37 AM, Venkatesh S <svenkat@yahoo-inc.com> wrote:
>
>> You could use ChainMapper and ChainReducer API in hadoop.
>>
>
>
>
>



Grant Mackey
UCF Research Assistant
Engineering III
Rm 238 Cubicle 1

----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.


Mime
View raw message