File APIs for Java Developers
Manipulate DOC, XLS, PPT, PDF and many others from your application.
The moose likes Hadoop and the fly likes Hadoop - One Map and many Reduces Big Moose Saloon
  Search | Java FAQ | Recent Topics | Flagged Topics | Hot Topics | Zero Replies
Register / Login
JavaRanch » Java Forums » Databases » Hadoop
Bookmark "Hadoop - One Map and many Reduces" Watch "Hadoop - One Map and many Reduces" New topic

Hadoop - One Map and many Reduces

Luan Cestari
Ranch Hand

Joined: Feb 07, 2010
Posts: 163

Hi Chuck Lam,

Suppose I have some data and I want process it iteratively grouping for a different key. I think this could be done by running some Hadoop Tasks, but each would have an initial load, that is the initial I/O and the mapping process.
My idea was a map once and then do several reduces. Those reduces would emit new maps for the next reduces.
Is it possible to do with Hadoop? What do you think about this approach?

Is there any planning to change the hadoop to run many reduces from the same map? ( I guess then way it is now it can't)

IMO, it look like that if you can do many reduces taking advantage of it is already in the memory would make the process faster than do many jobs with the same map.

Please, visit me for some cool tech post at
I agree. Here's the link:
subject: Hadoop - One Map and many Reduces
jQuery in Action, 3rd edition