You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 803 Next »

This effort is still a "work in progress". Please feel free to add comments.

Introduction

HAMA is a distributed framework on Hadoop for massive matrix and graph computations, currently being incubated as one of the incubator project by the Apache Software Foundation.

Goal

The Hama project goal is to provide easy matrix/graph computing programming environment on the Hadoop (distributed system). We are focusing on are as follows:

  • Compatibility
  • Scalability
  • Flexibility
  • Usability and Applicability

The overall architecture of HAMA

Below diagram is illustrates the overall architecture of HAMA.

               +--------------------------------------+
               |    Matrix/Graph Computing Program    |  User Applications
               +--------------------------------------+
           +------------------------------------------+
           |       HAMA : BSP, Angrapa, ..., etc      |  Computing Engines
           +------------------------------------------+ 
 +----------------------------------------------------+
 |           Hadoop : HDFS, HBase, ..., etc           |  Distributed Storage Systems
 +----------------------------------------------------+

BSP

The BSP package is a implementation of BSP over Hadoop RPC(sockets). By using a BSP model which is based on the concept of a superstep, during which processes perform computations using local data, a more rapid and sensitive program will be allowed.

The BSP package consists of the following components:

Matrix

Graph

Shell/DSL

  • Hama DSL (Domain Specific Language) in Groovy – Work in progress
  • Hama Shell – Work in progress
  • No labels