Grid-computing portals and security issues

  • Authors:

  • Ali Raza Butt;Sumalatha Adabala;Nirav H. Kapadia;Renato J. Figueiredo;José A. B. Fortes

  • Affiliations:

  • School of Electrical and Computer Engineering, Purdue University, 1285 Electrical Engineering Building, Mail Box 89, West Lafayette, IN;School of Electrical and Computer Engineering, Purdue University, 1285 Electrical Engineering Building, Mail Box 89, West Lafayette, IN;Capital One Services, Inc., Glen Allen, VA;Department of Electrical and Computer Engineering, University of Florida, Gainesville, FL;Department of Electrical and Computer Engineering, University of Florida, Gainesville, FL

  • Venue:
  • Journal of Parallel and Distributed Computing - Scalable web services and architecture
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

Computational grids provide computing power by sharing resourcesacross administrative domains. This sharing, coupled with the needto execute untrusted code from arbitrary users, introduces securityhazards. Grid environments are built on top of platforms thatcontrol access to resources within a single administrative domain,at the granularity of a user. In wide-area multidomain gridenvironments, the overhead of maintaining user accounts isprohibitive, and securing access to resources via useraccountability is impractical. Typically, these issues are handledby implementing checks that guarantee the safety of applications,so that they can run in shared user accounts. This work shows thatsafety checks--language-based, compile-time, link-time orload-time--currently implemented in most grid environments areeither inadequate or limit allowed grid users and applications. Asurvey of various grid systems is presented, highlighting theproblems and limitations of current grid environments. A runtimeprocess monitoring technique is also proposed. The approach allowssetting-up an execution environment that supports the fulllegitimate use allowed by the security policy of a shared resource.For shell-based applications, performance measurements of theproposed scheme show up to 2.14 times less overheads as compared tothe case where all applications including the shell aremonitored.