Skip to content

ravikanthreddy89/MR-MPI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

This is the MapReduce-MPI (MR-MPI) library.

Copyright (2009) Sandia Corporation.  Under the terms of Contract
DE-AC04-94AL85000 with Sandia Corporation, the U.S. Government retains
certain rights in this software.

----------------------------------------------------------------------

MapReduce is the operation popularized by Google for computing on
large distributed data sets.  See the Wikipedia entry on MapReduce for
an overview of what a MapReduce is.

The MR-MPI library is a simple, portable implementation of MapReduce
that runs on any serial desktop machine or large parallel machine
using MPI message passing.

As a user, you write a program which calls the MR-MPI library and you
provide functions that operate on your data such as a map() and a
reduce().  These functions are invoked by the library on single
processors, so that you typically do not need to write any parallel
code to perform a MapReduce.  You can also use a provided Python
wrapper on the library.  You can also use a provided script-language
interface to the library called OINK.

The library is written in C++ and can be called from C++ or from C or
other hi-level languages such as Fortran or a scripting language.  A
Python wrapper for the library is provided.  If you want to run on a
single processor, a dummy MPI library is provided to link against.  To
perform MapReduces in parallel, you need to link against an installed
MPI library.

The MR-MPI library and the accompanying software is licensed under the
Berkeley Software Distribution (BSD) License, which basically means it
can be used by anyone for any purpose.  See the LICENSE file in this
directory for details.

The most current version of the library including all bug fixes and
new featues can be downloaded at
www.sandia.gov/~sjplimp/download.html.

The authors of the library are Steve Plimpton and Karen Devine at
Sandia National Laboratories who can be contacted at sjplimp, kddevin
at sandia.gov.  Or see www.sandia.gov/~sjplimp.

This MR-MPI distribution includes the following files and directories:

README			   this file
LICENSE			   the BSD License
doc			   documentation
examples		   simple examples of MapReduce programs
mpistubs		   dummy MPI library
oink                       OINK scripting framework for MR-MPI library
oinkdoc                    documentation for OINK
python			   Python wrapper files on MR-MPI library
src			   library source files
user                       user-contributed MapReduce programs

Point your browser at doc/Manual.html to get started.