GRID Security Review - Computer Network Security - page 100

GRID Security Review

Lazaros Gymnopoulos 1 , Stelios Dritsas 2 ,
Stefanos Gritzalis 1 , and Costas Lambrinoudakis 1

1 Department of Information and Communication Systems Engineering
University of the Aegean, GR-832 00 Samos, Greece
{lazaros.gymnopoulos,sgritz,clam}@aegean.gr
2 Department of Informatics, Athens University of Economics and Business
76 Patission St., GR-104 34 Athens, Greece
[e-mail]

Abstract. A Computational GRID is a collection of heterogeneous computing
resources spread across multiple administrative domains, serving the task of
providing users with an easy access to these resources. Taking into account the
advances in the area of high-speed networking, but also the increased computa-
tional power of current micro-processors, Computational GRIDs or meta-
systems have gradually become more popular. However, together with the ad-
vantages that they exhibit they are also contributing to several problems associ-
ated with the design and implementation of a secure environment. The conven-
tional approach to security, that of enforcing a single, system-wide policy,
cannot be applied to large-scale distributed systems. This paper analyzes the se-
curity requirements of GRID Computing and reviews a number of security ar-
chitectures that have been proposed. Furthermore, these architectures are evalu-
ated in terms of addressing the major GRID security requirements that have
been identified.

1 Introduction

Distributed Computing is all about harnessing unused computing resources, from
across a computer network, so as to address workload challenges posed by demanding
computer applications. During the last decade, Distributed Computing has matured
from the notion of simple workload balancing to a ubiquitous solution that has been
embraced by some of the world's leading organizations across multiple industry sec-
tors.
While Distributed Computing harnessed the full potential of existing computer re-
sources by effectively matching the supply of processing cycles with the demand
created by applications, even more importantly it has paved the way for GRID Com-
puting. As I. Foster, C. Kesselman and S. Tuecke state: “GRID Computing has
emerged as an important new field, distinguished from conventional Distributed
Computing by its focus on large-scale resource sharing, innovative application, and in
some cases, high performance orientation” [1].
Although GRID Computing has, since mid 90's, gained the commercial, as well as
the scientific, interest, its exact definition is not yet clear. According to I. Foster and
C. Kesselman: “A Computational GRID is a hardware and software infrastructure that
provides dependable, consistent, pervasive and inexpensive access to high-end com-
putational capabilities” [2]. GRID Computing is concerned with coordinated resource