A storage area network (SAN) is a type of local area network that is designed to perform the task of transferring data for storage and retrieval and replication on networks using high-end servers. SAN is an architecture that attaches remote computer storage devices to a server. It also, supports fiber network technology. SAN technology uses a low-level network protocol to transfer disk blocks unlike Network attached storage (NAS) that employs file-based protocols; where computers request a part of an abstract file instead of a disk block.
Such devices work on Transport Control Protocol and Internet Protocol (Thornburgh & Barry, 2000). With the rapid change in technology, there is an increase in data rate; therefore, there is need for organizations to come up with ways of managing, protecting, and maintaining the information, while decreasing business risks and storage cost. IBM as a company has come up with a list of servers that can be employed, namely, blade servers, blue gene, system z mainframe, power systems and cluster servers.
Devices that provide higher storage capacities and fast access performance in a cost effective manner, and that can meet the rising demands for solutions in e-business, need clustering to allow clients to explore all the possible means of expanding and developing new services. This is aimed at providing high ease of use for computing in various fields such as finance, engineering or architecture and to deliver services that are beneficial to all at a reasonable price.
Clustering has been a contributing factor towards the rise of powerful computers; it is a technique of connecting several servers or multiprocessors to work on workloads as a single entity in a number of computing resources. The following are some of the benefits of clustering: Centralized system management—clusters allow for maintenance of distributed clusters. Greater processing capacity; since it uses multiple servers, the system can handle large amounts of data. Clustered systems allow for maximum utilization of resources. They share large unexpected amounts of processing power across a number of workloads.
Resource consolidation- one cluster can change the processing power of each workload and it can handle many tasks thus allowing for both resource utilization and resource consolidation. Cluster systems allow for disaster recovery; for instance, incase of failures the machines can restore it. It also, allows for geographic server consolidation Another server that can be incorporated that is easy to configure and manage is IBM System z. It applies an open mainframe technology, virtualization and IBM’s consolidation. It employs java and Linux.
System z facilitates all operations and applications in an easy and safe way at a reasonable cost. System z provides better results in I. T architectures that are safe, smart and flexible such systems impact positive changes in I. T field and Business. Thirdly, one of the servers that is less costly and allows one to save up to $ 12000 is the blade server; it is eleven times faster; it reduces energy costs by 90%, unlike other servers that generate a lot of energy. This enables one to manage energy in a period where there is a rising cost of energy. Also, there is a 65% reduction in connectivity cost.
Fewer cables are required during installation unlike other servers, thus blade server offers one, more choices and control at a cheaper price. Blade Center architectures have multiple layers that accommodate scalable and changing infrastructures. Moreover, it contains some important components like memory, processors as well as input and output devices that can be used in a variety of applications (IBM).
IBM retrieved on June 17, 2009 from: <http://www-03. ibm. com/systems/clusters/about/ > Thornburgh, R. H. & Barry J. S. (2000). Storage area networks: designing and implementing a mass storage system. Australia: Prentice Hall PTR.