Purpose-This paper aims to describe the problems faced by MyAIS, an open access repository of scholarly content, the reasons for moving to a cloud infrastructure and the design of the system. Design/methodology/approach-MyAIS is supported by an IBM HS22V multiple blade server using virtualization technology. A single blade server runs both MyAIS's back-end that handles the processing of digital repository and front-end that handles the web server. The whole system runs on a Linux platform. The virtualized servers deliver an infrastructure as a service (IaaS) to the Faculty of Computer Science, University of Malaya researchers as well as students. Findings-Sharing resources in the cloud cuts maintenance cost and leaves a smaller carbon footprint due to its modular design. Virtualization makes it easy to monitor access traffic and usage of hardware resources on-the-fly. The scalability and expansion of MyAIS is improved. It allows for faster system recovery in case of hard disk failure and speedy file move during migration. Larger bandwidth improves reliability and management of bottlenecks. Research limitations/implications-There are still bottlenecks at the wide area network level resulting in sluggishness during peak hours. Practical implications-There is significant improvement in storage expansion through shared hardware resources. Originality/value-Cloud as an IaaS helps reduce the total cost of ownership and maintenance, increase system scalability to handle increased traffic, meet increasing demand for storage space for full text resources, and reduce downtime. It makes backup and disaster recovery easier. It is possible to accommodate other university repositories in future.
Abstract-The numerical solution of a computationally intensive model becomes more complex in terms of execution time required by a single processor. To speedup the computation, a suitable parallel computing architecture is required. This paper attempts to achieve a fast finite difference solution of electric potential in an EHD ion-drag micropump. A 2D Poisson's equation is solved on a cluster of low cost computers using MATLAB. Numerical solution is obtained for the different mesh refinements and then the execution time, communication time, speedup and efficiency of parallel system are analyzed. The results showed that the speedup and efficiency of the system increases by increasing the grid points. The results also reveal that for each data size there is an optimum number of workers for obtaining the parallel numerical solution in minimum processing time.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.