Administrator's Guide: Introduction

From Gcube Wiki
Revision as of 14:14, 28 November 2007 by Vangelis (Talk | contribs)

Jump to: navigation, search

Purpose

Welcome to the gCube platform Administrators Manual. The purpose of this manual is to guide the first installation, management and daily administration of a grid infrastructure based on the gCube platform. gCube is a versatile, rich featured grid platform that has been developed in the DILIGENT European IST research project [ref]. The platform follows the Service Oriented paradigm and exploits and extends various other grid middlewares and collaborative tools like the Globus Toolkit 4 [ref], gLite [ref] and gridsphere portals [ref] in order to facilitate the sharing and collaboration of data in a form of grid computing based distributed digital library.

Being a mosaic of different services and tools is the reason why up to a point the installation of a new gCube infrastructure and the initialization of a Digital Library can be a daunting process. The current manual covers the basic steps of the core components installation like the DHN and Collective Layer services, provides an overview picture of the gCube architecture with emphasis to its security characteristics.

gCube Architecture Overview

gCube enables scientists to declaratively and dynamically build transient Virtual Research Environments --VREs-- by aggregating and deploying on-demand content resources, application services, and computing resources. It also monitors the shared resources during the VREs lifetime guaranteeing their optimal allocation and exploitation. Finally, it provides mechanisms to easily create VREs dedicated web portals through which scientists can access their content and services. The gCube system is realised as a service oriented framework composed of a set of interacting services providing:

  • support for the creation and operation of on-demand, transient virtual research environment;
  • features necessary for handling shared content and application resources;
  • access to information sources and applications provided by third-parties;
  • a set of typical DL functions, like search, annotation, personalisation, document visualisation.

These services are designed to exploit the gLite middleware and are capable to consume the high computational and storage capabilities of the Grid infrastructure released by the EGEE [ref] project. Thus gCube services support complex and time consuming functionalities, while focusing on optimizing resource usage and satisfying QoS contracts

From the logical point of view, the gCube system is organised according the following layers:

Collective Layer

The Collective Layer enhances existing Grid collective services with the functionalities able to support the complex services interactions required by the Digital Library Layer. The Collective Layer contains services that are not associated with any one specific resource but rather are global in nature and manage interactions across collections of resources.

Digital Library Layer

The Digital Library Layer selects, integrates and enhances a set of reliable and dependable production-quality services, developed in digital library projects and applications, in order to cover the fundamental functionalities required for any virtual research environment in the e-knowledge area. The services of this layer provide submission, indexing and discovery of mixed-media objects (documents, videos, images, environmental data, etc.), and the management and processing of these objects through annotation, composition, cooperative editing, etc. It also supports the dynamic creation and access to transient virtual research environment.

Application Specific Layer

The Application-Specific Layer contains application specific services. Third party providers are enabled to migrate their data or functional components to the gCube framework. Specifications are being produced that facilitate the plug-in of legacy components needed to support user-specific scenarios and enable the re-use of existing content and applications.

The current manual focuses mainly on the Collection Layer components which are the cornerstone of any gCube installation. The rest of the components are either deployed dynamically from the Keeper service or are application/user community specific (for instance the end user Portals)

Intended Readership

This manual targets system administrators and people responsible for setting up a gCube based Grid environment. It assumes fluency in Linux-based usage and administration, and knowledge of Globus Toolkit 4 Java WS-Core inner details. Since, some components take advantage of the gLite middleware, experience in setting up gLite-based grid infrastructures would be useful. Finally, as with all grid platform, also in gCube security plays a paramount role in the architecture. Both system administrator's and end users should be familiar with concepts like, digital certificates, certification authorities, authorization & authentication, etc.

Related Documents

Apart from this Administrators Manual, we have also made available two additional manuals: - the Users Manual [add link], which provides usage information and guideliness for the end-user of the two user communities that currently exploit the platform, namely ImPEct and ARTE. - the Developers Manual [add link], which target application programmers that they want to reused and extend the rich Service Oriented, grid API, provided by gCube.

Additional material that will help system administrators is the - gLite 3.0 Administrators Guide [provide link] - Globus Toolkit 4 Administrators Guide.

Regarding the architecture and inner details of gCube, the interested reader can visit the official gCube platform web site [1].

Lexical Conventions

Lexical conventions (if any, otherwise omit) used throughout the document.

Problem Reporting

Where to report problems about this manual (not the software).