Computer Networks Principles - Assignment Example

Computer Networks Principles of SECTION Introduction This paper will discuss two designs for two computer networks, the first design will discuss the issues regarding a LAN design and the second one will discuss the routing design and necessary configurations. The first scenario, discusses a small training business. As a systems and networking advisor, we need to design a computer network for the business based on the structure of the company and the number of computers and we will consider the fact that the network needs to be scalable and reliable. In the first section of this report, we will discuss the most important computer network design phases for an enterprise network and we will address suitable equipment recommendation where needed. In order to design or upgrade a computer network for business, we need to consider the goals for the network design and the possible problems that will exist in the network after the design is finished or the problems that already exists in the system. Network designer needs to cooperate with the client completely and find out their business and technical goals. For instance, the network designer should be aware of the new applications that client wants to place in the business environment. The designer needs to discover the amount of network availability which is needed to help the key sections of the business to be able to function correctly. By defining the business goals in an accurate way, a network designer starts a project that will finally lead to "success" in his/her job. We will discuss these issues in more detail below: Step 1: Identifying Business Goals/Needs Most of the times, the network design is accomplished by identifying the structure of the business. For instance, the logical design of a network is based on the different departments which exist in the network. So, it is a good start to discover the structure of the departments in the company as it is accomplished in the current assignment: "The business occupies a two-storey building and a small single storey supplementary building.The larger building has a business broadband connection using ADSL over ATM (PPPoA). The broadband connection has an 8Mbps download and 1 Mbps upload. The broadband connection terminates (POP) into the MDF, which is provided by the local Telecommunications Company. On the ground floor there are two training rooms - one containing 14 computers, the other containing 12 computers. The administration department is located at the centre of the ground floor and contains 10 computers plus a file/printer server. On the first floor there is a multimedia centre, which contains 16 computers, and a design/development department, which has 24 computers. The supplementary building houses the test department and contains 18 computers. This building is 500m away from the main building. It is company policy, wherever possible, to buy Cisco Networking Products." This information helps us to discover all main user communities and to make adequate network traffic flow. Common business goals of designing a LAN for businesses is to increase the revenue, increasing the performance of the employees, reducing networking costs, increase in the security and a better service for customer support. Step 2: Identifying Technical Goals/Needs After discovering the business goals, we need to identify the technical goals in current business too. The technical goals are often based on the business goals and sometimes, it is hard to discover the technical needs. The most important technical goal in order to design an enterprise LAN (or in a larger scale which could be WAN) in this assignment is scalability. There are other important factors too. For example availability of the network is another important issue to be considered. Scalability is considered to be the quality of a network to keep its functionality and performance considering the changes in network size or volume. The businesses today are rapidly growing and the number of users in a network changes in a short time. Networks need to be able to use various applications, opening new sections in their offices and so on. The design of network for the business case study is in a way that the increase of employees or the offices have a minimum impact on the quality and the performance expected from the network. All of the subnets are designed in way that more computers could be added. Also there is an estimation of adding three other rooms/offices to the company, each one capable of supporting thirty devices. The equal amount of traffic distribution by switches allows each part of the LAN to benefit from equal band-with with other LANS. Therefore it can be said that the order of the devices and equipments in the network is in a way that it makes the network scalable. There is another important factor which should be considered while designing a LAN and it is the availability of the network to its users. Knowing the amount of need for the network in hours, can have a great impact on the way the network needs to be designed. In a network that the possibility of being down is around an hour a day, we need to have a clear understanding on how much this network availability will influence the company. If the business will face with trouble and being online would be vital in each second, the design of the network should support the availability of the network. Backup systems are needed or the devices that can be replaced in the shortest possible time or the availability of using two or more ISP services. This way, if any malfunction happens, we can have more chances of being available and we can ensure our clients that the network is always accessible. This design costs more and needs a more complicated design. Considering the "small business" which is our case study in this assignment, and the scalability which has the first priority, the design in case of availability is considered to be a normal design. The design of the current network is shown in the next page. In this design we have assumed that each room/office/section which is connected to a switch is capable of having 30 IP address and the subnet masks are assumed to be based on 11111111.11111111.11111111.11100000. By this subnet mask, we are able to have 8 different subnets in our network each one with a capacity of 30 computers. Therefore we can consider two other rooms or two set of 30 computers to be added to our buildings in future too. This is considered due to the scalability policy of our network. Observing this LAN design diagram, we can easily find the most vital devices needed for our network and it can be helpful for future upgrades and enhancements of the devices and the business network. A LAN design diagram has many benefits, it gives us a better understanding of the physical and Logical Locations of the most important devices in a network, for instance, the locations of servers, computers, MDFs , Firewalls in addition to the modems, routers or hubs. Having enough information on the way the devices in a network are connected to each other, the network administrator can control the network and can better configure the important connected devices. For instance, the routing scheme in a mid-sized LAN could be determined by the location and special settings of layer 3 equipments. In our current design as you can see a multi-port router located at the ground floor is responsible for all the traffic in the buildings. A cable from the MDF comes to the router and the router is responsible for distributing equal traffic to the rooms. The design is so simple but it eases the upgrading options for the network since the main control device is the router. Fortunately, in order to upgrade or enhance the network, an administrator just needs to login to the device and form configuration changes which are needed. It is done by changing subnet masks and the ranges of IPs for each room and the total buildings. The issues discussed so far are mostly regarding the physical design of the network but solutions regarding scalability and availability are mainly discussed in the logical design of a network. There are some steps toward drawing a logical design for a LAN. The first step is to select a topology for a network. In our design, the network topology is star. A network topology demonstrates the configuration of the devices or the way they are arranged to perform the maximum scalability and availability of a network, also other important factors such as security. In this phase, the main network partitions are discovered and the way each device connects to other parts of the network is demonstrated. By defining a network topology, we can demonstrate the size of any network and we can simply define the scope of it. The first step is to create a layered and hierarchical design of the network devices. In a hierarchical model, a computer network is divided into three separate layers in a way that each layer could demonstrate some special functions. This way of network design is also called as modular design that simply enables network designers to expand the network design as it grows or needs to be changed. This way, the costs for the expansion or change in the network design would be a small subset of all the costs needed to design a new network. So, we can say that selecting a right topology and designing a hierarchical design of a network could highly help future designers and developers to make sure that the network is scalable. In order to draw a hierarchical network topology for our case study here in this assignment, we use three layers, as is demonstrated in the image below. The network in this design is described in a hierarchical topology with 3 layers, core, distribution and access layer. The core layer is mostly the start point of a network which as you can see, it may be a WAN or ISP which provides the total traffic for our enterprise network. The second layer can be a set of devices that distribute the received traffic from the layer one to all the switches in the buildings. As we have demonstrated at the physical design each room or section benefits from a switch that all the computer devices, servers or printers are connected to the router of that room. The current design is highly scalable and it distributes the network traffic equally to all the devices in the building, so due to the good download and upload speed received from the MDF to the router, we can say that each section receives a band-with of more than 1Mbs for download which is quite good. Core layer is also referred to as "the backbone" of a computer network. This name is used because it is the main gateway of a network for a building or a LAN with subnets and all the incoming and outgoing traffic is passed through a high-speed switching backbone and it needs to be designed in a way that data transmission happens as fast as possible. One of the main issues to be considered is this part of the network layer is that no device should be used in order to manipulate the sent and received packets. No packet filtering device because of slowing down the speed of the network. The important thing to remember is that none of the computer devices used by the user should not be a part of this layer and as you can see, the design of the network only shows a multi switched router which is connected to the core layer and this ensures that the traffic flow is not restricted and the speed remains as high as possible. The second layer in our network design is a router which functions as a access point between all the devices in the network and also between the access layer of the network and core. Some of the works that can be done on distribution layer includes the ability to divide departmental access of computers to the LAN or forming a workgroup for computers to be able to access to each other. Virtual LAN implementation is also possible and virtual LAN routing could be accomplished. One more important work that could be done on this layer is security issues. It is recommended to use a router that supports security check software and a firewall. Current layer in our design only includes a router which is the access point of the network and the network administrator/designer needs to be aware of the routing capabilities of the router. Routing functionality can be done statically or dynamically. Static routing is known as the easiest routing technique. In this way, network administrator needs to find routes and manage the way data is sent through the LAN. This way of routing in the network is much more reliable. Another advantage is that it consumes far less bandwidth for the transmission devices and it does not waste CPU cycles and far less device memory is consumed. Static routes can also be cheaper. There are many benefits while using a static routing system but it has some inherent limitations too. The most important disadvantage of these routers is that they are not manageable or if a device is replaced and the topology is changed, the network admin has to manually insert the necessary changes where is needed. Considering the time the network is growing and many routers are used in the network topology, it completely is not possible to configure the routing tables. Dynamic routing does not involve these problems and it is much more popular for mid-sized networks. In a business environment, the network design has to be done according to the needs of the enterprise regarding the scalability, availability and security. In our case study we have a limited amount of hardware options and we have to configure the network in a way that it could be cost effective and also it could be reliable. In some of the offices, we need to provide more band-with for multimedia applications purposes. In this situation and considering that we have currently the capacity to add two subnets to our network, we can give these subnet access to the places which are in more need for band-with, for example, half of the multimedia department could use one subnet and half of them could use the other subnet. SECTION 2: Introduction Section 2 of the assignment discusses the routing design and configuration of a wide area network. Below is the scenario according to the assignment: "Anglia International is an educational organisation with its main headquarters in London. Singapore and Tokyo are regional offices. The IP range of the whole network is a subnet of the IP address 144.69.0.0/16. The network connection to the ISP will use address 146.18.5.1/30. The LANs attached to routers London and Tokyo will each support 15 hosts. LAN0 and LAN1 attached to Singapore router must support 24 and 30 hosts respectively. Each LAN must be capable of operating at 100Mbps.The ISP connection will be a static default route. An access control list (ACL) would be applied so that on LAN 0 only hosts 0 to 7 are to be allowed access to all hosts within the Company Network and any Internet address. All other hosts on LAN 0 would be prevented from reaching any other host or network. All hosts on LAN1 and the hosts on the LANs connected to London and Tokyo routers are allowed to reach any address in the Company Network including the Internet address 192.168.10.1. Only web traffic is permitted at 192.168.10.1. The security administrator has discovered an employee misusing the email system by using a rogue Email server at 198.0.0.1. All hosts are to be banned from reaching this host for sending and retrieving email (SMTP/POP3)." 1. A critical assessment of two addressing schemes for the Company Network. In this section, two addressing scheme will be used for the network. The first design includes the division of network traffic from London's router to the three other sections of the network equally, including the LAN in London, the Tokyo router and the Singapore router. In this design equal number of IPs is given to each subnet. This design has some major problems. For instance, the development of the network would be a difficult work in future since re-configuring all IPs is needed in a case that a new office is required to be added in Berlin for example. Even if we want to assign some of the IP addresses of a subnet like London to the Berlin, this will provide the limit in band-with for Berlin. In the second IP addressing scheme, we assume that further offices would be added in future due to the development of the educational organisation and our focus would be on the scalability of the network. We will design the network in a way that the management and configuration of the network could be done in London. Address scheme Below is the range of IP addresses that allows 5 more subnets to be added to the current subnets in future. London London Router: 144.169.0.0/19 Subnet Mask: 144.169.244.0 Broadcast Address: 144.169.63.255 Host Range: 164.169.32.2 to 164.169.63.254 Subnet Address: 144.169.32.1 Tokyo Tokyo Router: 144.169.64.0/20 Subnet Mask: 144.169.240.0 Broadcast Address: 144.169.95.255 Host Range: 144.169.64.2 to 194.169.95.254 Subnet Address: 144.169.64.1 Singapore: Singapore Router: 144.169.128. 0/21 Subnet Mask: 144.169.248.0 Singapore LAN1: 144.169.136.0 Singapore LAN2: 144.169.144.0 Address scheme Below is the logical diagram of the complete network. In this diagram, the router located in London is also capable of supporting five more subnets in order to provide the scalability of the network in future. ACLs Below is the access lists for routers on the company network which configures them to implement the current policy: "An access control list (ACL) would be applied so that on LAN 0 only hosts 0 to 7 are to be allowed access to all hosts within the Company Network and any Internet address. All other hosts on LAN 0 would be prevented from reaching any other host or network. All hosts on LAN1 and the hosts on the LANs connected to London and Tokyo routers are allowed to reach any address in the Company Network including the Internet address 192.168.10.1. Only web traffic is permitted at 192.168.10.1." Router-SINGAPORE(config)#access-list 10 permit 144.169.136.1 144.169.136.8 Router-SINGAPORE(config)#access-list 10 deny 144.169.136.8 144.169.143.254 The first line states that hosts 0 to 7 from LAN0 which is 144.169.136.1, are allowed to send and receive data and the second line states that all other hosts on this LAN are not allowed to connect with any other host in the entire network. The next lines of router configuration commands will be written as follows: Router-SINGAPORE(config)#access-list 10 permit 144.169.144.1 144.169.176.254 Router-LONDON(config)#access-list 11 permit any Router-TOKYO(config)#access-list 11 permit any The first line shows that all the hosts in the LAN1 are located in Singapore are allowed to access to all information on the network. The second and third lines also state that all the hosts on routers located in London and Tokyo are allowed to access any packet which is sent or received through their routers. A description and comparison of Distance Vector and Link State routing Protocols "There are two main types of algorithms for IP routing: Distance Vector Routing and Link State Routing.Basically, Distance Vector protocols determine best path on how far the destination is, while Link State protocols are capable of using more sophisticated methods taking into consideration link variables, such as bandwidth, delay, reliability and load. Distance Vector protocols judge best path on how far it is. Distance can be hops or a combination of metrics calculated to represent a distance value. The IP Distance Vector routing protocols still in use today are: Routing Information Protocol (RIP v1 and v2) and Interior Gateway Routing Protocol. Distance-vector routing protocols are simple and efficient in small networks, and require little, if any management. However, they do not scale well, and have poor convergence properties, which has led to the development of more complex but more scalable link-state routing protocols for use in large networks. A Link-state routing is a concept used in routing of packet-switched networks in computer communications. Link-state routing works by having the routers tell every router on the network about its closest neighbours. The entire routing table is not distributed from any router, only the part of the table containing its neighbours. Some of the link-state routing protocols are the OSPF, IS-IS and EIGRP. Novell's NLSP (NetWare Link State Protocol) is also a link-state routing protocol, which only supports IPX. This type of routing protocol requires each router to maintain at least a partial map of the network. When a network link changes state (up to down, or vice versa), a notification, called a link state advertisement (LSA) is flooded throughout the network. All the routers note the change, and re-compute their routes accordingly. Link State Routing protocols provide greater flexibility and sophistication than the Distance Vector routing protocols. They reduce overall broadcast traffic and make better decisions about routing by taking characteristics such as bandwidth, delay, reliability, and load into consideration, instead of basing their decisions solely on distance or hop count."(Javvin.com, 2007) Due to the more benefits of Link State routing protocol, we suggest implementation of the network by this protocol. References Routing Protocols, Link State and Distance Vector Routing, viewed on December 13, 2007. Available from http://www.javvin.com/routing-protocols.html Kennedy Clark, Kevin Hamilton. 1999. CCIE Professional Development: Cisco Lan Switching. Cisco Press. Catherine Paquet. 1999. Building Cisco Remote Access Networks. Cisco Press. www.cisco.com www.cis.ohio-state.edu/hypertext/information/rfc.html Read More