Secondly, there is a rise in the need for quick decision making within the business environment. In times of business crisis, organizations use the information to make decisions. Therefore, there is need for data warehouses that can process information much faster than the centralized database systems (Taniar, Leung, Rahayu, & Wiley InterScienc, 2008). On this note, the research has majorly focused on technologies that can help to improve data processing speeds to achieve higher efficiency.The parallel database system is a new approach of data processing. It involves the use of large parallel processing computers that can achieve higher throughput as compared to one computer system. Parallel databases can be divided into various categories. First, there exists the shared memory architecture where the multiple processors are connected and share the main memory. The second category is the shared disk architecture in which each computer node has its main memory but they share the storage memory. in the recent past, the shared nothing architecture has come up and this is unique in that each node has its own main memory and storage memory. The hybrid architecture such as cluster structures have shared nothing and shared disks. In this technology a group of connected computers is used for data processing. These technologies have come up as a result of the demand for speed in processing big data (Taniar, Leung, Rahayu, & Wiley InterScienc, 2008). While these technologies were usually unavailable in the last few decades, they are now more common due to the rise in the demand for speed and the availability and big storage networks, making data processing a complex task for the serial processing databases.The cluster technology is one of the technologies used in parallel databases systems. In this method, a group of computers is interconnected, each with its main memory and processing capacities. Unlike in serial processing where information is processed by one computer, the cluster connection uses various processors to process the data. From another point of view, the various processors shared the load being processed. In data warehousing, information goes through the stage of loading, building indexes and evaluating queries. In a serial system, the above processes are done chronologically but by only one machine. On the other hand, parallel databases work together in all the three stages to provide a quick through
Taniar, D., Leung, C. H. C., Rahayu, J. W., & Wiley InterScienc, (2008).
High-performance parallel database processing and grid databases. Hoboken, N.J: Wiley.
Please type your essay title, choose your document type, enter your email and we send you essay samples