Scenario Proposal of a Server Cluster

The purpose of this document is to provide a proposal for a server cluster that will be introduced to Kucera Clothiers. This proposal is based upon a scenario of Kucera Clothiers is looking to improve communications between their corporate headquarters that is located in Chicago and its 3,500 other locations throughout the world. Kucera employs over 100,000 employees at their different locations throughout the world. Included within the proposal that will be proposed to Kucera Clothiers are some system requirements, costs, anticipated return on investment, security issues, and information management. This document is intended for anybody looking to gain a basic knowledge or understanding of how a project proposal can be introduced to a company.

Kucera Clothiers should maximize use of current resources in order to provide a network service that can accommodate any and all IT services today, but also well into the future. Kucera can maximize their return on investment of the existing network infrastructure and yet still provide scalable solutions for years to come. The use of a network server cluster can allow Kucera Clothiers to provide network services to all employees throughout the world with a 99.99% uptime guarantee. Providing services with a 99.99% reliability will greatly improve customer service, sales/customer relations, and efficiency throughout the company. The ability for employees to access data from anywhere at any time throughout the world is vital in today’s globalized markets.

A networked server cluster allows multiple servers to share valuable resources in order to handle any request, at anytime, from anyone. A server cluster functions by having multiple network interface cards (NICs) installed in multiple servers. One NIC allows users to access services and resources from the public network, however the 2nd NIC allows the different servers to communicate with each other and share valuable resources on a private network without interfering with any requests coming from the public network. Clustered servers can provide services such as: e-mail, web, file sharing, collaboration, database, and other specialized services. In the event that an additional service is needed, the service simply needs to be enabled on the cluster to allow the organization to grow with future needs or requirements. Server clusters provide much value to an organization by providing a high availability so that users in different time zones can access the servers as needed, thus allowing the company to maximize productivity. Due to the use of multiple servers, if one server or device should fail, the others can share the load until the device can be repaired or replaced (RedHat Inc., 2009). The design of a server cluster allows for the computing power to be located in one centralized location, in contrast to incurring high costs to locate the high performance computing out to each of the 3,500 locations worldwide. A server cluster can provide much value to an organization by providing all services from one centralized location with high performance servers that can easily handle the processing workload of all the company’s employees from around the world while maintaining a 99.99% uptime guarantee to allow the users to be more productive.

Prior to installing a server cluster there are some requirements that need to be met so that the cluster can easily mesh with the organizations requirements. There needs to be enough server rack space to accommodate the cluster. There also needs to be enough power provided to the server racks to meet all the power needs of the system. The location where the cluster is going to be housed needs to be an environmentally controlled climate. The facility that the cluster will be located needs to have diversified network paths so that in the even one should go down; other paths are able to handle the network traffic. The facility where the cluster is going to be located also needs a managed network switch with enough available ports to handle the configuration of multiple virtual LANs (VLANs) in order to support the cluster. The location where the servers will reside needs to be a physically secure location. When supporting any system it is important to ensure all of the system requirements can be met before obtaining the system.

The cost estimate of the system includes all necessary peripherals to include hardware, software, licensing, and support. The total cost of the system is $22,235. This cost is for the purchase of 5 Dell PowerEdge 1950 III servers with Redhat Linux installed on the systems and a 3-year basic on-site support. The exact system options are listed below (Dell, 2009):
PowerEdge 1950 III Quad Core Intel® Xeon® E5405, 2x6MB Cache, 2.0GHz, 1333MHz FSB
Additional Processors Single Processor only
Memory 32GB 667MHz (8x4GB), Dual Ranked DIMMs
Operating System Red Hat Enterprise Linux 5.2, 2S, FI x64, 1yr, Auto-Entitle, Lic & Media
Multipack Operating System None
Operating System Addition None
OS Partitions None
Chassis Configuration Rack Chassis with Static RapidRails
Backplane 1×2 Backplane for 3.5-inch Hard Drives and SATA Optical
Primary Controller SAS 6/iR Integrated Controller
2nd Controller and HBAs None
Hard Drive Configuration Integrated SAS/SATA No RAID, SAS 6/iR Integrated
Primary Hard Drive 1TB 7.2K RPM Universal SATA 3Gbps 3.5-in HotPlug Hard Drive
2nd Hard Drive Discounted Upgrade! 1TB 7.2K RPM Universal SATA 3Gbps 3.5-in HotPlug Hard D
3rd Hard Drive None
4th Hard Drive None
PCI Riser Riser with 2 PCIe Slots
Power Supply Energy Smart Redundant Power Supply with Y-Cord
Power Cords Power Cord, NEMA 5-15P to C14, 15 amp, wall plug, 10 feet / 3 meter

Utilizing current network infrastructure, facility environment, and networking devices, the total cost of ownership for the server cluster can be identified by the total cost of all necessary hardware and support.

The return on investment for the system can be immediately identified by the design of the system. Even though the initial cost of the system may seem high to managers who are unaware of IT costs, the total cost of the system is miniscule compared to the cost if the processing power was decentralized and located at each individual location. An additional return on investment can be clearly seen because the centralized location of the servers ensures the servers are physically secured and the servers are centrally managed to ensure the data on the servers is secured. Also due to the servers being centrally located the need for IT support personnel at each of the locations is non-existent. All IT support will be provided in one centralized location, which saves the company money from paying for support at multiple locations and varying fees. The scalability of the system provides much added value to grow as the company grows. The server cluster provides an immediate return on investment by cutting IT support costs, improving overall system and data security, and being scalable to grow as the company demands.

In conclusion, a centralized server cluster can provide much value to Kucera Clothiers by allowing all employees to have access to valuable data at anytime from anywhere. The cluster provides necessary IT services today, but it is also scalable to meet the IT needs of the future too. The server cluster will provide an immediate return on investment by cutting IT costs at multiple locations by housing the high performance servers in one centralized location. Due to the centralized design of the system, all security concerns of the systems and their data will be easily addressed. By using current resources that are available, the total cost of the system includes nothing more than the hardware and support costs. Kucera Clothiers wishes to improve communications throughout the organization and this server cluster provides the means to do so.

1. | Red Hat Cluster Suite. Retrieved May 16, 2009, from | The World’s Open Source Leader Web site:
2. The Dell Online Store: Build Your System. Retrieved May 16, 2009, from Dell Laptops, Desktop Computers, Monitors, Printers & PC Accessories Web site: