Analyzing Technical Requirements

This chapter is from the book

MCSE Training Guide (70-221): Windows 2000 Network Infrastructure Design, 2nd Edition

Learn More Buy

This chapter is from the book

This chapter is from the book 

MCSE Training Guide (70-221): Windows 2000 Network Infrastructure Design, 2nd Edition

Learn More Buy

Apply Your Knowledge

Exercises

Exercise 2.1: Creating an IP Addressing Scheme

This exercise demonstrates the process of creating an IP addressing scheme that can be implemented across an enterprise. You will complete Table 2.1 to create the addressing standards for the Lighthouse, Inc. corporate network.

Estimated Time: 30 minutes

Lighthouse Inc. has corporate headquarters in Bangor, Maine and four remote sites in Nashua, N.H.; Lowell, Mass.; Providence, R.I.; and Buffalo, N.Y. Each site should adhere to the same IP addressing standards. In each location, there need to be at least eight client PCs, as well as four routers and/or switches, five servers, and five printers. You should allow for the addition of one or two of each type of device to the network to accommodate some minor anticipated growth over the next three years. There is also the possibility that one or two additional locations might be added to the network within three years. Those sites will need to adhere to the same standards within the selected IP address space.

The address space of 192.168.16.0 with a subnet mask of 255.255.255.224 has been selected for the enterprise.

Table 2.1 Addressing Standards for Lighthouse, Inc.

Answers to Exercise 2.1

When you have completed this exercise, you should have a completed table that looks like Table 2.2.

Table 2.2 Answers to Addressing Standards for Lighthouse, Inc.

This solution meets the company's growth requirements by allowing for the following in each location (see Table 2.3).

Table 2.3 Growth Requirements

The solution also allows for the addition of at least two additional sites that can follow this addressing scheme.

Exericse 2.2: Performing a Hardware Assessment

This exercise demonstrates the process of assessing the existing hardware in a company against the goals of the new network infrastructure design. Complete Table 2.4 based on the information provided regarding the design of the Golden Gardens, Inc. network infrastructure.

Estimated Time: 15 minutes

Golden Gardens, Inc. is a small company with one location. In its offices, it has installed a network infrastructure with four file servers running Microsoft Windows NT 4.0 and 16 client workstations. Six of the workstation computers have 486 DX 100Mhz CPUs with 16MB of RAM, six have Pentium 233Mhz CPUs with 32MB of RAM, and four have Pentium II 266Mhz CPUs with 64MB of RAM.

Management at Golden Gardens, Inc. has mandated that part of the new infrastructure design include the rollout of a new application that has recently been purchased. The minimum CPU and memory requirements for this application are a 200Mhz CPU and 64MB of RAM. You want to perform an assessment of the existing hardware so that you can make recommendations to company management regarding any hardware upgrades that must be performed in order to support the new application.

Complete Table 2.4 in order to summarize your results and present your recommendations to Golden Gardens, Inc.'s management.

Table 2.4 Results Summary

Answers to Exercise 2.2

When you have completed this exercise, you should have a table that looks like Table 2.5.

Table 2.5 Answers to Results Summary

Exercise 2.3: Analyzing Client Computer Access Requirements

In this exercise you will compile a list of client computer access requirements for use in your infrastructure design. We will use a fictitious company, MedEx, for this exercise.

Estimated Time: 10 minutes

MedEx sells medical equipment and supplies to various hospitals in the northeast. They are headquartered in Friendship, Maine, where they have a staff of 100 people. Approximately 30 salespeople with laptops are responsible for visiting local hospitals and taking orders.

These orders are taken using a custom-built client application. The client application collects the sales orders, and then the salesperson connects to headquarters to transmit the orders to a database located there.

Several executives located in the headquarters run large queries against the sales database in order to analyze sales information and perform growth projection. Also, purchasers run queries against the sales database in order to forecast future sales and generate orders for more products from suppliers.

List the factors in determining the requirements for the following groups:

• Sales personnel

• Executives

• Purchasers

Answers to Exercise 2.3

• Sales personnel: The sales staff will need dial-up connectivity from various locations in order to transmit their sales orders.

• Executives: They will need large amounts of bandwidth in order to process their queries against the sales database.

• Purchasers: They will need large amounts of bandwidth in order to process their queries against the sales database.

Exercise 2.4: Analyzing the Existing Disaster Recovery Strategy for Client Computers, Servers, and the Network

In this exercise you will analyze the disaster recovery strategies and plans implemented by a company in order to ensure that your network design accommodates them. In this exercise we'll use a fictitious company called NorthWind Publishing.

Estimated Time: 10 minutes

NorthWind is a publishing company located in New York. They have their headquarters there, which holds 300 people, a mainframe, and 20 servers. Two of the servers at the headquarters hold databases that are mission-critical to the operation of the business. They have a number of other satellite offices in other cities. These offices have between 5 and 50 employees. The larger offices have a server, but the smaller offices do not. Central company functions, such as HR and finance, are done at headquarters, and the company data is stored on the servers. The satellite offices have writers, graphic artists, and administrative employees who create the books, articles, and other various documents.

Based on the preceding information, create a list of disaster recovery concerns.

Answers to Exercise 2.4

• Are the servers at headquarters and the satellite offices being backed up?

• Is there offsite storage for tape backups?

• Is there hardware fault tolerance for the servers?

• Is there failover capability for the mission-critical servers? Is the data on the server being replicated to another site?

• In the offices without servers, are client computers being backed up? If so, where are those tapes stored?

Review Questions

1. What components comprise the total user population for an organization?

2. What is the purpose of the gap analysis?

3. What are two typical network topology models?

4. What is net available bandwidth?

5. In what way does latency affect the performance of applications on a network?

6. What are some areas you should consider when developing a security strategy for the network infrastructure?

7. What are some of the typical wiring types used in today's network infrastructures?

8. What is the function of a repeater?

9. What is the purpose of an IP address?

10. Why is it unwise to combine the deployment of a new application or an upgrade to an existing application with the implementation of a new network infrastructure design?

Exam Questions

The first five exam questions refer to the Case Study presented earlier in this chapter.

1. Assuming that you decide to recommend a centralized approach to Dewey, Winnem, & Howe's network structure, what type of topology would you recommend?

   1. Hub and spoke

   2. Partially meshed

   3. Fully meshed

   4. Flat loop

2. What impact does the outsourcing of computer support have on the size of the end-user community within the law firm?

   1. Outsourcing has no impact on the size of the end-user community.

   2. Outsourcing reduces the size of the end-user community.

   3. Outsourcing increases the size of the end-user community.

   4. Outsourcing helps maintain consistency and the size of the end-user community.

3. What challenges are posed by Dewey, Winnem, & Howe's lack of centralized administration over TCP/IP address assignment in the remote offices?

   1. IP addresses might be duplicated.

   2. IP addresses might be assigned by a DHCP.

   3. Certain hosts won't need IP addresses.

   4. IP addresses may be assigned that are outside the scope of the company's IP addressing scheme.

4. What issues surround Dewey, Winnem, & Howe's use of proprietary database applications?

   1. High MTTR

   2. High MBTF

   3. End-user access to database resources

   4. Network latency

5. What remedies might you consider in your new infrastructure design to solve the problem of users complaining of long delays when trying to the access the system in the corporate headquarters?

   1. Decrease the end users' workload by reducing the number of applications that are run on each workstation.

   2. Purchase hardware with a lower MTTR.

   3. Consider upgrading the hubs on each floor of the headquarters building with network switches.

   4. Consider revising the company's TCP/IP addressing scheme.

6. You have been hired by the New Shoes, Inc. organization to create a new design for its network infrastructure. New Shoes is in the process of acquiring one of its competitors, Best Foot, Inc. Company management has stated that the new network infrastructure must be as decentralized as possible. Given this decentralized approach, how should you plan for resource distribution in the new infrastructure design?

   1. Resources should be physically located in the New Shoes corporate headquarters.

   2. Resources should be physically located in the Best Foot corporate headquarters.

   3. Resources should be distributed and physically located near the WAN links that connect the two companies.

   4. Resources should be distributed throughout the enterprise, physically located near the end users who take advantage of them.

7. You are responsible for network design and administration for your company, New Riders, Inc. The company has just informed you that it will be opening a new office in Gnome, Alaska. It wants you to establish a T1 line to connect that office as soon as possible. What issues might you face in providing this connectivity?

   1. Inclement weather might delay the installation of the T1 lines.

   2. T1 connectivity might not be available in that particular location.

   3. A T1 link might provide more bandwidth than is necessary for that particular office.

   4. A T1 line to Gnome, Alaska would be too expensive.

8. You are responsible for network design for West Coast Importers, Inc. Your manager, Pedro, comes to you trying to understand the difference between capacity and net available bandwidth. What do you tell him?

   1. Net available bandwidth is always greater than network capacity.

   2. Net available bandwidth is always less than network capacity.

   3. Net available bandwidth and network capacity are the same and therefore are always equal.

   4. Net available bandwidth equals network capa- city minus protocol overhead and bandwidth used by protocols that carry nonuser data.

9. You are the network manager for a company named Saint Paul Purchasing. The management at Saint Paul Purchasing is very upset with you this week because the network had an availability rating of 99.70 percent. They are upset because 99.70 percent availability is totally unacceptable given the nature of their business and the requirements that are placed on the network infrastructure. Given that the network was available for 99.70 percent of the time over the course of one week, for how long was the network unavailable?

   1. 5 minutes

   2. 30 minutes

   3. 1 hour

   4. 3 hours

10. Harbor Lights, Inc. has contracted with you to design a network infrastructure. The company would like to ensure that everyone can access its data servers and email servers most efficiently. What are some typical data and system access patterns that you might expect to find while conducting your investigation of Harbor Lights?

    1. Few users actually access the servers located at corporate headquarters.

    2. Users tend to access the servers most frequently in the morning and the evening.

    3. Users access the email servers at the same frequencies all day long.

    4. The file and email servers are overworked.

11. You're designing a network infrastructure for a company named Jones Services, Inc. Management at Jones is questioning your recommendation for the use of private IP addresses. They ask you to justify your recommendation. Which of the following are benefits associated with the use of private IP addresses?

    1. Private IP addresses allow you to connect your network to the Internet.

    2. Private IP addresses give you a much broader range of subnet and host addresses.

    3. Private IP addresses are centrally managed, registered, and assigned by Internet service providers.

    4. Private IP addresses help minimize the impact on your overall internetwork should you decide to change Internet service providers.

12. The management at Good Foods, Inc. has hired you to create a new network infrastructure designed for its corporate network. They want you to focus your security efforts on protecting the internal network from external intruders. You tell company management that although this is a valid concern, there are other areas of security upon which you will need to focus your efforts. What are some of the other areas of security that will need to be addressed while you create your network infrastructure design?

    1. Physical security

    2. Telephone security

    3. Internal access security

    4. Administrative security

13. You're in the process of creating a network infrastructure design for the Sun Dial Breads Company. You develop an IP addressing scheme using private IP addresses. Sun Dial management wants you to explain the nature of an effective IP addressing scheme so that they can understand your undertaking. What are some of the elements of an effective IP addressing scheme?

    1. Hierarchical

    2. Random

    3. Private

    4. Meaningful

14. You're in the process of designing a network infrastructure for Laurel Foods, Inc. Laurel Foods is currently using Windows NT 4.0 servers to offer file and print services. It uses Internet Information Server 4.0 as its Web server platform. The client workstations in the corporate headquarters access the Windows NT file servers for data storage. Client workstations also access the NT file servers for office productivity applications, including the Microsoft Office suite and Lotus Notes applications. The clients also use a 32/70 emulator to connect to a mainframe for corporate financial data. The company plans to open two new branches across town. It is your job to design the connectivity for the two new offices. What issues must you address in order to provide adequate connectivity for the users in the new offices?

    1. Legacy protocols must be routed to the new locations in order to give users access to the mainframe applications.

    2. You must install frame relay circuits to the remote locations in order to provide adequate bandwidth for users to run the office productivity applications.

    3. You must use private TCP/IP addresses in order to connect the remote locations to the corporate headquarters.

    4. Connectivity must be provided 24 hours a day at each location in order for users to accomplish their jobs.

15. General Services, Inc. has contracted with you to develop a new network infrastructure designed for its enterprise. General Services will be acquiring one of its competitors, T.H. Howard and Sons, and will incorporate the T.H. Howard network into its own infrastructure. T.H. Howard has several remote locations, each with many end users. Responsibility for supporting the existing General Services network, as well as the planned merged network, will fall on the General Services three-person technical support team. Consider the impact of the three-person technical support team on the design of the overall network. In what way will this affect your design?

    1. Your network design will have to incorporate a new network support team.

    2. Your network design will have to incorporate remote support tools and methods to allow this existing support team to do its job effectively.

    3. Your network design will have to include a reduction in services offered across the enterprise.

    4. Your network design will not be affected by the existing technical support structure.

Answers to Review Questions

1. When determining an organization's total user population size, consider not only its employees, but also any contractors who work on the company premises or at remote sites. In addition, you must consider any vendors, partners, or clients who will be accessing the local internal network. Any user who accesses resources of the internal corporate network must be considered in the total user population for that organization. See "Analyze Company Size and User and Resource Distribution."

2. Gap analysis allows you to determine the differences between the existing environment and the environmental goals proposed by company management. The gap analysis will form the basis of your network infrastructure design and will help you decide which features and improvements will need to be included in your design. See "Evaluating Technical Environment and Goals."

3. Two common topology models found in internetwork design are the hub and spoke model and the mesh model. In a hub and spoke network, one side is selected as the central point of communication, and all the other sites connect to this central point. The links from the hub to the remote sites form the "spokes," and the entire architecture looks like a wheel. In a mesh topology, there is no central hub. Each site is connected to one or more of the other sites. In a fully meshed topology, each site is connected to all the other sites on the network. In a partially meshed topology, each site is connected to one or more sites, but not necessarily to all the other sites. See "Assessing Available Connectivity."

4. Net available bandwidth is the amount of bandwidth available to end users for transmission of user data. This figure is arrived at by determining the overall available bandwidth of a network link and subtracting from it the bandwidth used by protocol overhead and other information traversing the link that does not carry user data. See "Assessing Bandwidth and Latency."

5. Latency refers to the amount of time between the moment when a network station is ready to transmit data and the moment when the transmission is completed successfully. Increases in latency can be the cause of application timeouts, which would cause the application to cease functioning. Loss of functionality, also referred to as downtime, can be very expensive. See "Assessing Bandwidth and Latency."

6. Some areas of concern while developing a security strategy for your network infrastructure design include physical security, internal access security, and external access security. See "Security Considerations."

7. Some typical wiring types in today's network infrastructures include unshielded twisted pair, shielded twisted pair, coaxial cable, and fiber-optic cable. WAN devices often use serial cables such as the EIA/TIA 232. See "Network Infrastructure, Protocols, and Hosts."

8. A repeater is a device placed on the network to extend the distance that a network may span. Data enters the repeater on one side and is amplified and sent out on the other. The use of a repeater helps restore signal strength that is diminished over long cable runs. See "Network Infrastructure, Protocols, and Hosts."

9. An IP address is a number that uniquely identifies each end station connected to the network. Each station must have an IP address, and that IP address must be unique across the network. No two stations can have the same IP address. See "The IP Addressing Scheme."

10. Should issues arise during implementation, the combination of projects such as these helps complicate the troubleshooting process. See "Upgrades and Rollouts."

Answers to Exam Questions

1. A. If you decide to implement a centralized approach to the network infrastructure for Dewey, Winnem, & Howe, you should recommend a hub and spoke topology. With a hub and spoke network topology, you would use the Manhattan headquarters as the hub site, and all connectivity to the remote sites would connect from the center of the hub. This provides connectivity from each site of the enterprise to the others through the Manhattan headquarters, which becomes the central point of control. See "Assessing Available Connectivity."

2. C. Outsourcing of computer support by the various remote offices increases the overall size of the end-user community because the firms that are charged with supporting the computer systems for these remote offices will be utilizing network resources belonging to the firm. These additional users, although they aren't company employees, will place additional demands on the network infrastructure. Support employees from the outsourcing vendor may work on-site within the law firm's remote offices or may work at their own corporate headquarters with some sort of remote connectivity established from their offices to the Dewey, Winnem, & Howe internal network. If this is the case, an additional burden will be placed on the infrastructure, and you will need to investigate the type of connectivity that is established, as well as the access patterns of the support employees themselves. See "Assessing Available Connectivity."

3. A, D. The IP addresses are assigned by individuals in each remote location without being managed by a central authority. Therefore, there is a likelihood that duplicate IP addresses will be assigned to workstations throughout the enterprise. In addition, it is very likely that a local administrator will begin to assign IP addresses that are not within the scope of the company's selected addressing scheme. See "TCP/IP Infrastructure."

4. C. The company's use of proprietary database applications leads to the concern of end-user access to database resources. You'll need to ensure that database resources are accessible to all users who need them, whether they are in corporate headquarters, in a remote location, or accessing the network via dial-up or VPN services. Answers A and B refer to hardware performance, not to software, so they are incorrect. Answer D is also incorrect because, although it's true that latency is a concern for application performance, it is a concern for all applications, not specifically for proprietary database applications. See "Analyzing Client Computer Access Requirements."

5. C. The company is using hubs on each floor of the corporate headquarters, so all users are sharing the total bandwidth available. Replacing the hubs with network switches will potentially allow you to provide the full Ethernet bandwidth to each end station. This will dramatically reduce the overall load on the network and will reduce the delays experienced by end users while they access the network. See "Network Infrastructure, Protocols, and Hosts."

6. D. Management wants the network to be decentralized. This will require you to distribute network resources throughout the enterprise physically located near the end users who take advantage of them. Each location will have an administrator who is responsible for managing each resource. The scope of that administrator's responsibilities will be limited to the single resource or few resources over which he has control. See "Security Considerations."

7. B. When designing connectivity between various physical work sites, you need to investigate the availability of communication between those locations. It might be easy to assume that certain types of connectivity are available in highly populated areas, such as busy city or suburban neighborhoods. However, certain types of communication facilities might not be available in rural areas, mountainous areas, or other sparsely populated locations. An investigation of the types of communications available will allow you to understand the options available to you before you create your design. See "Assessing Available Connectivity."

8. D. Net available bandwidth refers to the total bandwidth available to end users on that particular link. You arrive at this figure by starting with the capacity of that network link and subtracting bandwidth used by protocols that do not carry end-user data. You will also want to subtract any protocol overhead associated with the protocol you've selected to carry that user data. The figure you are left with is the bandwidth that is actually available to each end user to carry data across the link. See "Assessing Bandwidth and Latency."

9. B. An availability rating of 99.70 percent over the course of one week results in a downtime of approximately 30 minutes. In many environments, 30 minutes downtime per week is unacceptable, especially if that downtime is experienced all at once. When measuring network availability in percentages, the 99 on the left side of the decimal point is assumed. The two digits following the decimal point represent the acceptability or unacceptability of performance and availability. Tolerances for downtime will vary from one organization to another, depending on the nature of the business organization and the requirements placed on the network infrastructure. 99.99 percent availability is always strived for but rarely achieved. See "Performance, Scalability, and Availability."

10. B, C. In many organizations, you might expect to find typical data and system access patterns. One such pattern is that of user access to file servers. This access tends to be most frequent first thing in morning, when users arrive for work and log into the system. Access to email servers tends to be consistent throughout the course of the day. See "End-User Usage Patterns."

11. B, D. Private IP addresses allow you much greater flexibility and a much wider range of subnets and host addresses compared to what is available in the remaining public address space. Private IP addresses also minimize the impact on your enterprise network should you choose to change Internet service providers. The public address assigned to you by your ISP will be configured on relatively few devices on your internetwork. Should you change ISPs and receive a new publicly assigned address, you need only change that address on those few devices. The remaining devices in your enterprise will remain the same. See "The IP Addressing Scheme."

12. A, C. In addition to focusing external access security to protect the internal network from external intruders, you need to focus on other areas of security while creating your network infrastructure design. These other areas include internal access security and physical security. See "Security Considerations."

13. A, D. An effective IP addressing scheme is one that is hierarchical in nature. An IP addressing scheme should also be meaningful, with some basis in the architecture of the company itself. Public and private addressing space can be effective in an IP addressing scheme, provided that they are applied in a hierarchical and meaningful fashion. See "The IP Addressing Scheme."

14. B, C. To provide for this scenario as stated, legacy protocols will not need to be transmitted across the network, but some form of connectivity will be required, although nothing in the scenario requires 24-hour access. In this case, answers B and C are the most likely. See "Evaluating Technical Environment and Goals."

15. B. Your new network infrastructure design must incorporate remote support tools and other methods to help the three-person technical support team effectively provide support across the entire organization. You can safely assume that this team will be overworked, but you might ease the support burden by building in the capability to support network services remotely and effectively. See "Technical Support Structure."

Suggested Readings and Resources

1. Adam, Kelli, and Rob Scrimger. MCSE Training Guide: TCP/IP, Second Edition. Indianapolis, IN: New Riders Publishing, 1998. (ISBN: 1562059203)

2. Berg, Glenn. Network+ Exam, Fast Track Network+ Series. Indianapolis, IN: New Riders Publishing, 1999. (ISBN: 0735709041)

3. Cisco Systems. Cisco IOS 12.0 Switching Services. Indianapolis, IN: Cisco Press, 1999. (ISBN: 1578701570)

4. Ratliff, Randy. Network+ Certification Guide. Indianapolis, IN: New Riders Publishing, 1999. (ISBN: 073570077X)