Introduction and requirements for Giraffa - GiraffaFS/giraffa GitHub Wiki

Introduction

Giraffa is a distributed, highly scalable file system, which is aimed to be able to:

  • Support millions of concurrent clients
  • Store trillions of objects
  • Maintain Exabyte total storage capacity

Giraffa is utilizing features of HDFS and HBase in order to scale both the data storage and its metadata. Giraffa keeps directories, files, and blocks in a single HBase table distributed across Region Servers, while file data are stored in blocks on HDFS DataNodes.

Requirements

In order to use and develop Giraffa, you will need to install Java 7 (Java 8 has not been tested).

Giraffa is using Gradle 2.5 as a build system:

Giraffa currently runs using hbase-1.0.1 and hadoop-2.5.1. The links below are Apache mirror links to places to download them.

Like Hadoop, Giraffa uses Google Protocol Buffers to generate some intermediate java file.
Please refer to Protobuf Git for installation guidance.

It is recommended you run Giraffa on a Linux distribution. Windows with Cygwin, and Mac, are also working.

Optional Extras

It is highly recommended that you use Eclipse or InteliJ to work on Giraffa because it comes with Gradle and Git, which you will need in order to build Giraffa and to contribute source code.

If you choose not to use Eclipse, you may need to download Git separately. Follow git directions for installation.

⚠️ **GitHub.com Fallback** ⚠️