The computational grid offers services for efficiently scheduling jobs on the grid, but for grid-enabled applications where data handling is a most relevant part, the data grid kicks in. It typically builds on the concept of files, sites and file transfers between sites. These use a data transfer service, plus a replica manager to keep track of where replicas are located. The authors consider a multi-site, grid-aware data warehouse, which is a large distributed repository sharing a schema and data concerning scientific or business domains. Differently from typical grid scenarios, the data warehouse is not simply a set of files and accesses to individual files. It is a single distributed schema and both localized and distributed computations must be managed over that schema. Given this difference, it is important to study approaches for placement and computation over the grid data warehouse and this is our contribution in this book chapter.
Introduction And Motivation
Grid computing has the potential to provide users “on-demand” access to large amounts of computing power, just as power grids provide users with consistent, pervasive, dependable and transparent access to electricity, irrespective of its source (Baker et al., 2002). Simulation in industry can potentially benefit from this as computing power can be an issue in the time taken to get results from a simulation (Robinson, 2005a). This is further supported by the observation that the use of grid computing in scientific simulation has certainly proved beneficial in disciplines such as particle physics, climatology, astrophysics and medicine, among others. Thus, our first motivation is to inform simulation users in industry as to how the practice of simulation can benefit from grid computing.
Another motivation is the low adoption rate of grid computing outside of academic and research domains. At present a major proportion of grid users comprise of researchers (physicists, biologists, climatologists, etc. who are the primary stakeholder of the applications running on the grid) and computer specialists with programming skills (the providers of IT support to the stakeholders). This is not unexpected as the majority of applications using grid computing are research applications. The adoption of grid computing technologies by employees in industry has so far been relatively modest. One important reason for this is, although the employees are experts in their own discipline they generally do not have the necessary technical skills that are required to work with present generation grid technologies. A possible means to increase adoption is to incorporate grid support in software applications that require non-trivial amounts of computation power and which are used by the end-users to perform their day-to-day jobs. The commercial, off-the-shelf (COTS) Simulation Packages (CSPs) used in industry to model simulations are an ideal candidate for such type of integration. This chapter, thus, focuses on leveraging the practice of CSP-based simulation in industry through use of grid computing. Figure 1 shows the motivations of this research.
The remainder of this chapter is organised as follows. The second section gives an overview of the practice of simulation in industry and the CSPs used to model such simulations. The following two sections are devoted to grid computing and desktop grid computing respectively. The CSP-grid integration approaches are proposed in the subsequent section, followed by two CSP-grid integration case studies. The last section presents a summary of this research and brings the chapter to a close.Top
Simulation Practice In Industry
Defining Computer Simulation
A computer simulation uses the power of computers to conduct experiments with models that represent systems of interest (Pidd, 2004). Experimenting with the computer model enables us to know more about the system under scrutiny and to evaluate various strategies for the operation of the system (Shannon, 1998). Computer simulations are generally used for experimentation as they are cheaper than building (and discarding) real systems; they assist in the identification of problems in the underlying system and allow testing of different scenarios in an attempt to resolve them; allow faster than real-time experimentation; provide a means to depict the behaviour of systems under development; involve lower costs compared to experimenting with real systems; facilitate the replication of experiments; and provide a safe environment for studying dangerous situations like combat scenarios, natural disasters and evacuation strategies (Brooks et al., 2001; Pidd, 2004).