Creating and Managing Grid Services

There are many definitions of what constitutes a “Grid ” – the most common of which identifies a Computational Grids as: “a collection of distributed computing resources available over local or wide area networks-- and which appear to an end user as one large virtual computing system”. A significant effort has been invested by the Grid community into building software systems that may be used to manage such a virtual computing system. The Grid approach is an important development for both the business and scientific community [Laszewski03], and promotes a vision for sophisticated scientific and business collaborations. Until recently most of this effort was focused on managing resource ensembles, and co-ordinating access to such ensembles in a secure and efficient manner. The focus of this work was primarily driven by the need to handle differences between different kinds of hardware architectures (Intel vs. Sun vs. SGI), operating systems (Solaris vs. Win2000 vs. IRIX) and data repositories (structured databases vs. filesystems). To establish a ``virtual organisation'', it is also necessary to allow individual systems administrators-- managing one or more resource ensembles-- to exercise their individual authority. Each of these systems administrators would have control over their resources, and would not allow some centralised system to override access rights that they have defined for external users coming to their system. The predominant interest within existing Grid system