1. Why is it important to use a structured, systematic method for designing networks? What problems can occur if such methods are not used?
2. Why is it important to explore divisional and group structures of an organization when starting a network design project?
3. Discuss the term “scalability.” What does it mean? Why is it an important network design goal? What is some challenges designers face when designing for scalability?
4. A network design customer has a goal of 99.80 percent uptime. How much downtime will be permitted in hours per week? How much downtime will be permitted in
minutes per day and seconds per hour? Which values are acceptable in which circumstances?
5. What is a discontiguous network? Why does it pose challenges for a network designer?
6.Why is it important to characterize a network’s logical topology and not just its physical topology? What information does a logical topology illustrate that might be missed with just a physical topology?
7. Many basic networking books state that there are three network topologies, a bus, a ring, and a star. Based on the knowledge you gained in this chapter, explain why these simple topologies may not meet the needs of modern networks.
8. Compare and contrast full- and partial-mesh topologies. What advantages does a partial mesh have compared to a full mesh? What advantages does a full mesh havecompared to a partial mesh?
9. Why is it difficult to achieve true load balancing (as opposed to load sharing) in most networks?
10. What does physical security mean? What are some facets of physical security?
11.The following network numbers are defined at a branch office. Can they be summarized, and if so, what is the network number and prefix length?
12. What is a discontiguous subnet? Why do some enterprise networks have discontiguous subnets? Why don’t classful routing protocols support discontiguous subnets?
13. Discuss the configuration of IP addresses. When manually configuring IP addressing on a host, what parameters should you configure? If you suspect that there might be a problem with the addressing information that you entered, what commands can you use to troubleshoot? What are some disadvantages of manual configuration? What other options are there?
14. What are the most important criteria for selecting an internetworking device?
15. What is the difference between single-mode and multimode fiber? Which is faster?
16. Why are QoS features often necessary in LAN switches and routers?
17. What are the most important criteria for selecting a WAN service provider?
18. Why are QoS features often necessary in WAN routers?
19. List and describe five typical goals for a network design testing project.
20. What are some advantages of using industry tests for your network design testing project? What are some disadvantages of using industry tests?
21. What are some advantages of testing a new network design component on a production network during normal business hours? What are some disadvantages of taking this approach?
22. What is regression testing? Why is regression testing usually an important part of a network testing project? In what cases is regression testing not necessary?
23. How does multicast routing differ from unicast routing?
24. This chapter mentions serialization delay (also sometimes called transmission delay). What other types of delay are there? List and describe three other types of delay.
25. What are some techniques for reducing serialization delay?
1. Using structured network designs has two chief goals. First objective is that they build systems exploration is a more precise represent consumers’ requirements. Second objective was to produce manageable project by separating it into components that can be maintained as well as changed easily (Yeganeh et al., 2013). Failure to use these methods can result to several problems for the designing of networks, for instance; lack attention on data types, data flow, as well as procedure that change or access data.
2. The top-down networks designing procedure comprises exploring divisional as well as group configurations to get the individuals for whom networks will offer services as well as from whom one is supposed to get valuable info for the design to succeed (Yeganeh et al., 2013). The top-down networks designs are as well iterative. It is significant to get a comprehensive opinion of a client's requirements first to avoid being bogged down in information too quickly. Later, more details can be collected on technology preferences, scalability requirements, and protocol behavior. Top-down network designing appreciates that the physical design and the logical model can vary as more detail is gathered.
Scalability is the capacity of a network, process, or a system to manage an increasing volume of work in a proficient way or its capacity to be engorged to manage that change. In case of an enterprise, scalability signify that one is ready to accommodate a growing number of clients, customers, and/or consumers.
Management is in a position to experience the necessity of making huge amount of information available for employees, prospective customers, customers, suppliers and vendors (Yeganeh et al., 2013).
Scalability openly a?ects the key components or drivers of a company like Distribution, Production, Infrastructure and Networks. Scalability is an essential network designing objective for it is an ITSM (information technology service management) principles of design, delivery, create, management and support that covers the ITIL (Information Technology Information Library) framework (Yeganeh et al., 2013).
- Delay as well as Delay Variations
- Developers can underestimate the growing rate of a business and may offer a budget which, is underrated to grow a measurable system
4. Weekly : 22 hours
Daily: 33 minutes
Hourly: 36 seconds
5. Discontiguous Network
A discontigous network is a situation where one network have either two or even more classful sub networks linked together by diverse classful networks (Carofiglio et al., 2012). Discontigous subnets usually poses a major problem when incase of classful routing protocol or when trying to acquire intelligent summarization through the network.
- Illustrates the structural designs of networks
- Shows procedures of connecting gadgets in an orderly shape
- Helps in analyzing things that might affect scalability
- Helps the capacity to advance the network
The information in a logical topology that might be missed out when not well thought-out are: availability, adaptability, policies, network performance and routing as well as switching problems.
7. The bus, star, and ring topologies do not lend themselves well to redundancy as mesh topologies does. Their performance is also less compared to that of mesh topologies since lack any link delay among any two positions.
8. Partial and Full Mesh Topology
Partial and full mesh topology have got the following comparisons:
- Information flows from one router to another router in both topologies.
- When any component fail to function, data transfer is not affected.
Partial and full mesh topology has the following contrast or differences:
- Partial mesh topology has fewer links as compared to full mesh topology.
- Full mesh topology’s every switch or router connected to every single switch or router while partial mesh has less connections.
Full mesh topology offers good redundancy than provided by partial mesh topology
9. It is challenging because a routers might cache the interfaces that they utilize in case of destination hosts or even full destination networks, which implies that every traffic to the destination will spend the same period.
10. Physical security
Physical security means the safety of hardware, personnel, data, software and networks from physical events as well as actions which, could cause critical loss or damage to an agency, institution, or enterprises (Yeganeh et al., 2013). This includes safeguarding from fire, natural disasters, flood, theft, burglary, terrorism and vandalism. Physical security is mainly characterized by various aspects of physical security like, protective barriers, architecture and security guards or any other worker.
11. Prefix length is an easier method of articulating the subnet masks (Yeganeh et al., 2013). The length is symbolized by the amount of bits put in subnet masks; for example, if a certain mask is 255.255.255.0, it has 24 1â€™s according to the second description of the subnet mask, therefore the prefix length of the mask will be 24 bits.
12. Discontiguous subnets are means of travelling from a certain subnet to the other through same networks one would require to go thru subnets on the diverse network (Yeganeh et al., 2013). Classful routing procedures do not back discontiguous subnets for they summarize courses to only a single network that they are on. The discontiguous subnets are not in a position access other networks.
13. In the process of configuring IP addresses manually, one configures the default gateway as well as everything else that will continually be linked to that network (Yeganeh et al., 2013). In case of a challenge when addressing information one should use the ping and ipconfig command to see whether it is linked. tracert command is used to find where it can be held up or even where the hop stopped so that one can be able to get where linking stops.
- Manual configuration will only happen when several different systems are manually configured.
14. Criteria for choosing internetworking devices comprise the following (Carofiglio et al., 2012):
- Processing speeds
- Amount of memories
- Amount of dormancy introduced when devices transmits data
- Cells per second for ATM (or amount in packets per second)
- WAN and LAN technologies supported
15. Multimode fiber have got a comparatively larger light transporting core, regularly 62.5 microns or bigger in terms of diameter (Yeganeh et al., 2013). Single-mode fiber contains smaller light ferrying core of about 8 - 10 microns as diameter. Due to the narrow diameter in Single Mode Fiber, through which only a single mode will broadcast usually 1310nm or 1550nm, it carries greater bandwidth than the one carried by multimode fiber. However it involves light sources with slender spectral widths. Multimode fiber provides a higher bandwidth at higher speeds for medium distances.
16. QoS (Quality of Service) is a component of switches and routers that arranges traffic in terms of importance (Carofiglio et al., 2012). QoS improves the performance for important network traffic. It is suitable with LANs or VoIP phones with higher volume of local traffics.
17. It is important to for the following criteria first when decide on the WAN services provider:
Criteria 1 Network Transparency
Criteria 2 Interoperability with Application Performance Monitoring Tools
Criteria 3 Partnership with Application Sellers
18. QoS assist network managers to first identify traffic through types of service and shape traffic depending on bandwidth available (Carofiglio et al., 2012). QoS-enabled networks essentially place traffic into different buckets that resemble CoS (classes of service). This entails deciding which traffic to get first priority down to the traffic with lesser importance, according to preset policies or business needs.
19. Common technical goals include:
Improve network security. This is redesigning or improving the safety of a business’s network.
Upgrading network performance. Upgrading the performance of a network through application of new networks or upgrading existing network.
Increase network availability. It is achieved by implementing network redundancy structures.
Streamline network management. This is redesigning of the network management procedures.
Increase network scalability. The requirements for network in an organization change over time.
Industry tests can be used for simple networking.
Industry tests might be suitable in convincing customers to use certain network designs or devices.
Designers are able to ensure that networks are running appropriately during working hours. In case of any malfunctions he/she might troubleshoot them as well as repairing them (Carofiglio et al., 2012).
When network fail it could take some hours before network backup is gotten, this cause a loss of working hours.
22. Regression testing: It is a method of testing changes happening to PC programs with an aim of ensuring that previous programs are working with the changes (Carofiglio et al., 2012). It ensure the new system does not conflict at the stage of applications and components that are recognized for performing and working to a certain step before installation of the system. It does not check fresh features and upgrades
23. A Unicast transmission/stream convey IP packets to a single receiver on a system while Multicast stream convey IP packets to a group of hosts on a system (Carofiglio et al., 2012).
24. Propagation Delay: It states the time taken by each bit to shift from one particular end of the link to another, when an optical signal or electrical is sited onto the chain (Carofiglio et al., 2012).
Queuing Delay: When packets have to wait for other packets to be delivered, they experiences queuing delay.
Forwarding Delay: Denotes the processing period between the time, frame is entirely-received, and the time when packets have been put in an production queue,
25. At WAN end on the WAN link with less or equal bandwidth than E1/T1, it is advisable to activate both RTP/TCP header compression as well as LFI. These increase WAN link use and decrease serialization delay (Yeganeh et al., 2013). Since Layer 2 payload compressions are CPU-intensive, they are only recommended when they can be hardware-assisted or hardware-based.
Carofiglio, G., Gallo, M., & Muscariello, L. (2012, March). ICP: Design and evaluation of an interest control protocol for content-centric networking. In Computer Communications Workshops (INFOCOM WKSHPS), 2012 IEEE Conference on(pp. 304-309). IEEE.
Yeganeh, S. H., Tootoonchian, A., & Ganjali, Y. (2013). On scalability of software-defined networking. IEEE Communications Magazine, 51(2), 136-141.