Comparison between Different Networking Architectures Using OSPF, RIP and EIGRP

 Abstract

This report shows the comparison of different architectures using OSPF, RIP and EIGRP. My findings for this project are that OSPF has the highest traffic sent compared to RIP and EIGRP. This is because OSPF uses the shortest path and it consumes less bandwidth than RIP and EIGRP. The overall findings for the convergence activity throughout the 5 topologies is that EIGRP is the fastest protocol because it has the least delay time from failing and recovering time which are set at 240 seconds and 500 seconds. To summarise EIGRP is the best protocol no matter what topology is used, but the disadvantage of using it is that there is slow convergence and limited scalability. In today’s world OSPF is used and RIP is not the best protocol to use because the performance is very slow. I have achieved my aims and I am happy but there is still more configuration in which I could do such as looking at the delay and number of hops in a topology. Lastly, I would use OSPF because it can build a hierarchical and scalable network and it supports traffic engineering. Also it has many other benefits such as it is easy to configure, easy to manage and it allows routers to calculate routes that satisfy the incoming request.

Acknowledgements

I would like to thank my parents for the motivation and my supervisor Athanasios Tsokanos for the advice.

Contents page

Abstract 2

Acknowledgements 2

Contents 3

List of Figures 4

1. Introduction 5
   1. Aims 5
   2. Report structure 5
   3. Literature review 6
2. Background Research 6
   1. Wide Area Network (WAN) 6
   2. Multiprotocol label switching (MPLS) 6
   3. Resource reservation protocol (RSVP) 7
   4. Open shortest path first (OSPF) 8
   5. Routing information protocol (RIP) 8
   6. Enhanced interior gateway routing protocol (EIGRP) 9
   7. Automatic bandwidth 9
   8. Networking topology 10
   9. Cisco routers 10
   10.         Juniper routers 11
   11.         Network cards and ports 11
3. Methodology 11
4. Design 12
   1. Star topology 12
   2. Ring topology 12
   3. Tree topology 13
   4. Bus topology 14
   5. Full Mesh topology 14
   6. Partial Mesh topology 15
   7. Line topology 16
5. Implementation 16
6. Testing and Results 22
   1. Comparison 1 22
   2. Comparison 2 23
   3. Comparison 3 24
   4. Comparison 4 25
   5. Comparison 5 26

7. Evaluation 27
   1. Critical analysis of results 27
   2. Discussion 27
   3. Conclusion 27
   4. Successes and failures 28
   5. Management of the project 28
   6. Future improvements 28
8. References 29

List of Figures

Figure 1: Example of a wide area network 6

Figure 2: Example of a MPLS network 7

Figure 3: Example of a RSVP  7

Figure 4: Example of an OSPF 8

Figure 5: Example of a RIP  8

Figure 6: Example of an EIGRP  9

Figure 7: Example of auto-bandwidth  9

Figure 8: Example of different network topologies 10

Figure 9: Comparison table of Cisco routers 10

Figure 10: Comparison table of Juniper T-Series routers 11

Figure 11: Comparison of network cards and ports 11

Figure 12: Star Topology 12

Figure 13: Ring Topology 13

Figure 14: Tree topology 13

Figure 15: Bus topology 14

Figure 16: Full Mesh topology 15

Figure 17: Partial Mesh topology 15

Figure 18: Line topology 16

Figure 19: Application configuration 17

Figure 20: Profile configuration 17

Figure 21: Failure recovery configuration 18

Figure 22: Global attributes IP routing 18

Figure 23: Simulation stop for OSPF, RIP and EIGRP and run simulation for 20 minutes 19

Figure 24: OSPF DES statistics 19

Figure 25: RIP DES statistics 20

Figure 26: EIGRP DES statistics 20

Figure 27: OSPF parameters 21

Figure 28: RIP parameters 21

Figure 29: Overlaid Traffic Sent (bits/sec) on Star Topology 22

Figure 30: Overlaid Convergence Activity on Star Topology 22

Figure 31: Overlaid Traffic Sent (bits/sec) on Ring Topology 23

Figure 32: Overlaid Convergence Activity on Ring Topology 23

Figure 33: Overlaid Traffic Sent (bits/sec) on Mesh Topology 24

Figure 34: Overlaid Convergence Activity on Mesh Topology 24

Figure 35: Overlaid Traffic Sent (bits/sec) on Line Topology 25

Figure 36: Overlaid Convergence Activity on Line Topology 25

Figure 37: Overlaid Traffic Sent (bits/sec) on Partial Mesh Topology 26

Figure 38: Overlaid Convergence Activity on Partial Mesh Topology 26

1. Introduction

Implementing and configuring a network without much experience can be a difficult task, although I have some knowledge whilst studying a previous module called Network Protocols and Architectures. My project is about showing the comparison between different networking architectures using OSPF, RIP and EIGRP. This is an interesting project because it touches upon networking which is a major topic in Computer Science. I have always wanted to create a network using a network simulator and this is my dream to do so. This project will allow me to develop my understanding and improve my practical skills using Riverbed Modeler Academic Edition 17.5. First, I will research which hardware is best for my network and which cabling to use, then I will design the network and do the configuration to see which networking architecture is faster and efficient within a network. The key tasks that are required as part of the project are literature review, background research, methodology, design, implementation, testing, evaluation and gathering all of this into one final report.

1.1 Aims

The aims of this project are to:

• Research hardware and cabling needed for a network
• Describe the network that will be created
• Implement the network topologies
• Compare different networking architectures
• Make sure the configuration works using OSPF, RIP and EIGRP
• Test the speed of different networking architectures
• Create graphs to show the results of the comparison

1.2 Report structure

Chapter 1 introduces the report, shows the aims of the project and the literature review.

Chapter 2 details the background research undertaken of certain aspects of the project.

Chapter 3 focuses on what methodology was used and how the research was carried out.

Chapter 4 looks at the design of the network and different networking architectures.

Chapter 5 focuses on the configuration of the different networks and how the networks are implemented.

Chapter 6 is the stage where the networks are tested to see which one is the best in terms of speed; the evidence will be given in the form of graphs and tables where it will show the comparison.

Chapter 7 analyses the results obtained from the tests, critical analysis, improvements made, conclusion and what are my future plans with the project.

Chapter 8 is the list of references used throughout this project formatted using the Harvard style.

1.3 Literature review

The tasks which I have undertaken in order to research about this project are to review various websites, articles, journals, books and use my own knowledge using Riverbed Modeler. (Rani and Goyal, 2017) stated that MPLS offers the solution to the problems of traffic engineering, for example speed, network congestion and delay. MPLS sends data through labels attached to every packet; these labels are allocated between all the nodes making the network. In conclusion routers in MPLS take a small amount of processing time in forwarding the packets, also implementing of MPLS with traffic engineering decreases the congestion in the network. Finally MPLS suffers minimum delay and provides high throughput in comparison to conventional IP networks.

(Pana and Put, 2017) presents a simulation study of RSVP and its scalability extensions in OPNET Modeler. The results collected are put in context of the previous work done on the scalability of RSVP. It is found that the Summary Refresh extension significantly improves the performance of the protocol equally in terms of message processing and message generation. An essential finding of the paper is that the scalability of RSVP is not directly linked to the specifications of the protocol and its extensions, but to a certain extent to the design choices of the actual implementation under evaluation.

2. Background research

2.1 Wide area network (WAN)

A wide area network (WAN) is a geographically distributed private telecommunications network that interconnects many local area networks (LANs). In an enterprise, a WAN might consist of connections to a company’s headquarters, cloud services and other facilities. Usually, a router is used to connect a LAN to a WAN. Enterprise WANs let users to share access to applications, services and other centrally located resources. This is a benefit because there is no need to install the same application server, firewall or other resource in multiple locations. There are different types of WAN connections and they can include wired and wireless technologies. Wired WAN services can include multiprotocol label switching, T1s, Carrier Ethernet and commercial broadband internet links. Wireless WAN technologies can include cellular data networks like 4G LTE, as well as public Wi-Fi or satellite networks.

Figure 1 –Example of a wide area network

2.2 Multiprotocol label switching (MPLS)

This provides high performance packet control and forwarding mechanism for routing the packets in the data networks. It has changed into an important technology for efficiently operating and managing IP networks because of its greater capabilities in providing traffic engineering (TE) and virtual private network (VPN) services. It is not a replacement for the IP but it is an extension for IP architecture with including new functionalities and applications. The most important functionality is to connect a short fixed-label to the packets that enter into MPLS domain. Label is placed between Layer2 (Data Link Layer) and Layer3 (Network Layer) of the packet to form Layer 2.5 label switched network on layer 2 switching functionality without layer 3 IP routing. Also the packets in the MPLS network are forwarded based on these labels. MPLS networks have many benefits and these are there is enhanced performance, easy management, simple, secure, scalable and flexible.

Figure 2 – Example of a MPLS network

2.3 Resource reservation protocol (RSVP)

This is a signalling protocol that handles bandwidth allocation and true traffic engineering across an MPLS network. Like LDP, RSVP uses discovery messages and advertisements to exchange LSP path information between all hosts. On the other hand, RSVP includes a set of features that control the flow of traffic through an MPLS network. While LDP is restricted to using the configured IGP’s shortest path as the transit path through the network, RSVP uses a combination of the Constrained Shortest Path First (CSPF) algorithm and Explicit Route Objects (EROs) to decide how traffic is routed all the way through the network. Basic RSVP sessions are recognised in exactly the same way that LDP sessions are recognised. By configuring both MPLS and RSVP on the appropriate transit interfaces, it enables the exchange of RSVP packets and the establishment of LSPs. However, RSVP also allows configure link authentication, explicit LSP paths and link colouring.

Figure 3 – Example of a RSVP

2.4 Open shortest path first (OSPF)

OSPF introduces another layer of hierarchy into routing by allowing a domain to be partitioned into areas. This defines that a router within a domain doesn’t essentially need to know how to arrive at every network within that domain; it might be enough for it to know how to get to the right area. As a result, there is a decrease in the amount of information that should be transmitted to and stored in each node. Also OSPF allows many routes to the same destination to be assigned the same cost and this will cause traffic to be dispersed equally over those routers.

 

 

 

 

 

 

 

 

 

 

 

Figure 4 – Example of an OSPF

2.5 Routing information protocol (RIP)

A Routing information protocol (RIP) is one of a family of IP Routing protocols and it is an Interior Gateway Protocol (IGP) intended to distribute routing information inside an Autonomous System (AS). RIP is a simple vector routing protocol with lots of existing implementations in the field. In a vector routing protocol, the routers exchange network reachability information with their nearest neighbours. For example, the routers communicate to each other the sets of destinations that they can reach and the next hop address to which data should be sent in order to reach those destinations.

Figure 5 – Example of a RIP

2.6 Enhanced interior gateway routing protocol (EIGRP)

EIGRP is a network protocol that allows routers to exchange information more efficiently than with earlier network protocols. EIGRP came from IGRP (Interior Gateway Routing Protocol) and routers using both EIGRP and IGRP can operate because the metric used with one protocol can be translated into the metrics of the other protocol. Also EIGRP can be used not only for Internet Protocol (IP) networks but also for Novell NetWare networks and AppleTalk.

Figure 6 – Example of an EIGRP

2.7 Automatic bandwidth

Automatic bandwidth allocation allows an MPLS tunnel to automatically alter its bandwidth allocation based on the volume of traffic flowing through the tunnel. The new bandwidth is determined by inspecting the traffic flowing through the LSP. The user can configure an LSP with minimal bandwidth. Also the user can change the LSPs bandwidth allocation based on existing traffic patterns. The bandwidth adjustments don’t disrupt traffic flow through the tunnel. At the end of the automatic bandwidth adjustment time interval, the current maximum average bandwidth usage is compared with the allocated bandwidth for the LSP. When the LSP needs more bandwidth, an effort is made to set up a new path where bandwidth is equivalent to the current maximum average usage. When the attempt is successful, the LSPs traffic is routed through the new path and the old path is removed. When the attempt fails, the LSP continues to use its present pathway.

 

 

 

 

 

 

 

 

 

 

Figure 7 – Example of auto-bandwidth

2.8 Networking topology

A network topology is a topological structure of a computer network which can be physical or logical. The physical topology shows the location of different computer network essentials such as computers, cables and other devices. The logical topology visually displays network data flows from one device to another device. From time to time logical and physical topologies can be similar. There are five main types of network topologies and they are Star, Ring, Mesh, Bus and Fully Connected.

 

 

 

 

 

 

 

 

 

Figure 8 – Example of different network topologies

 

2.9 Cisco routers

There are many routers in which I have looked at to use for my design, I have researched the Cisco website and have compared the routers with each other to see which one would be the best for my network.

Figure 9 – Comparison table of Cisco routers

2.10 Juniper routers

Similarly there are many Juniper routers in which I have looked at to use for my design, I have researched the Juniper website and various other websites and have compared the routers with each other to see which one would be the best for my network. I have chosen to use the T-series routers because they are designed for high-end and core networks.

Figure 10 – Comparison table of Juniper T-Series routers

2.11 Network cards and ports

Here I have researched about the cables and legends because this is very important within a network. For example a CAT5e cable has transmission speeds of 1000mbps which will be good for a network compared to CAT6 cable which might not work well if there is interference and distance issues. But there is a disadvantage with CAT5e cable and that is the data sent is very limited and this can cause an issue with the network. Also copper cabling is good only to cover a short distance such as 100m with a data rate of 100mb/s, whereas fibre optic cable covers 500m with a data rate of 10 gb/s so this is much better.

Figure 11 – Comparison of network cards and ports

3. Methodology

I conducted secondary research for this project because this is vital for me to increase my understanding on networking architectures and OSPF, RIP and EIGRP. I grasped all of the knowledge from the internet and by using books and articles to form my own project. This was a tough process in which I have learnt a lot during the time. Also I used qualitative and quantitative approaches for this project. Firstly I did a literature review of some articles and case studies and then the centre of attention was the situation of the problem and how it can be solved. I used Riverbed Modeler to configure the network topologies to get statistical results to compare which one is fast and efficient to use in a network. Also I analysed the results that were the solution for the problem statement of the project. For this a network is designed and results are collected to show the comparison between different networking architecture using OSPF, RIP and EIGRP.

4. Design

4.1 Star topology

I created a star topology using the Juniper routers T640 because they are high performance routers and also they have the ability to scale well beyond their capacity. The connection in which I used to connect the five routers to the central point is 100baseT, this is because it has got good speed and it is very reliable. Also I have configured the routing protocol RIP that’s why the legend is on the bottom left corner. I have also got the application definition which specifies the number of applications to be configured and I have got the failure recovery which is to control the time and status of objects in the model. The profile definition is used to create user profiles. These user profiles can then be specified on different routers in the network to generate application layer traffic. There are many advantages of using this topology and they are that each router has its own cable and does not need to share the line with any other router. Also if a cable to a router is broken down, then only that router is affected. The rest of the network can run normally. New links to the central router can be without difficulty added, or existing links can be separated without affecting the rest of the network. It is also easy to discover a faulty cable because it will only affect a single link. A disadvantage of using this topology is that if the central router fails then the whole network will stop operating, in addition more cabling is needed for this topology compared to bus, tree and a ring topology. I have achieved to learn about what routers and cables are used in a star topology and how to configure the application and profile definitions.

Figure 12 – Star Topology

4.2 Ring topology

I created a ring topology which consisted of 10 Juniper T640 routers and the connection I used to connect all of them was 100baseT. Also I added the application and profile definitions because these are very important. In this topology traffic can be one-way around the loop or it can travel in both directions if a double loop is used. This ring network is quick; on the other hand if one router on the ring breaks down or a part of the network is disconnected, the network will not work. The main general example of the ring topology is the IBM Token Ring Network. An advantage of using this topology is that a single cable is used to connect all nodes, therefore many cables are not required making the speed of the network fast. Some disadvantages of this topology are that if for example the ring is broken, or the cable is faulty, then the whole network will come to an end to work. Also the ring must be broken to add new routers or remove existing ones, so this is another factor to take into account when choosing this topology. I have achieved to learn about what routers and cables are used in a ring topology, the advantages and disadvantages of using this topology and how to configure the application and profile definitions.

Figure 13 – Ring Topology

4.3 Tree topology

This topology is confusing although there are many benefits of using this topology because it integrates the features from bus and star topologies. I built this tree topology with 85 Juniper T640 routers and the connection used to connect all of them was 100baseT. Also I added the application and profile definition because these are vital in a network. There are some advantages when using this topology the first one is that if a star and bus topology can’t be implemented, this topology is a good alternative. Also this topology can be expanded into many levels which can be easily done. Another benefit of using this topology is that it is divided into separate segments so it makes it easy to maintain and manage. Each part is provided with dedicated point-to-point cabling to the central hub, so even if one segment is damaged the other segments would not get affected. Some disadvantages of using this topology are that because it is very basic and it relies a lot on the main bus cable, if this gets damaged then the entire network will be down. Also whenever more routers are added to the network it is hard to maintain the network and this is when it becomes complicated. I have achieved to learn about what routers and cables are used in a tree topology, the advantages and disadvantages of using this topology and how to configure the application and profile definitions.

Figure 14 – Tree topology

4.4 Bus topology

I built this bus topology using 18 ethcoax workstations and the link for all the workstations was eth coax. This topology uses a single cable to connect all of the computers in the network to each other. This network topology was adopted originally because running a single cable through all the computers in the network is more efficient and uses a smaller amount of wiring than other topologies such as tree and mesh. As before bus topology networks used huge coaxial cables, now it has changed. Some of the advantages of using this topology are that it is easy to implement and extend, also it is designed for temporary networks and primarily is cheaper than star, ring and tree topologies. There are quite a few disadvantages of using this topology and they are although this topology can handle large networks, this can be a warning because all of the computers are connected to the same main cable, therefore increasing the number of workstations in the network will have a big impact on the networks speed and quality. Also when more workstations are added onto the network, more data will be required to be transmitted and because all of the workstations are connected with the same main cable, all of the data has to travel down this single cable. There is an effect on this which is called data collision, this will make the network slower and prevent it from growing too big with bus topology. I have achieved to learn about what workstations and links are used in a bus topology and the advantages and disadvantages of using this topology.

Figure 15 – Bus topology

4.5 Full Mesh topology

I built this full mesh topology with 10 Juniper T640 routers and the connection used to connect all of them was 100baseT. Also I added the application and profile definition because these are vital in a network. A full mesh topology is when every router has to be interconnected with each other. This is used to connect devices without having to transmit or switch. The routing technique or the flooding technique is the design of how the full mesh topology works in a network. The message from the starting place is relayed through a path by simply going through router after router until it identifies the matched receiver. Through the use of routing method all of the routers are ensured that it is accessible upon transmission even if some of the routers are not working, this procedure is called the self-healing algorithm which makes it easier for the network engineers to troubleshoot issues as regards to the data spread. There are many advantages of using this topology the first one being a broken router wouldn’t divert the transmission of data in a full mesh network. Each and every router is connected to various other routers which make it easier to pass on data. A broken device will be ignored by the signals and will then find a new one that is connected with the router. Also extra devices in a full mesh topology will not affect the network connection that’s why it will improve the traffic in the network. This topology can also handle high amount of network traffic as every additional device into the network is considered a router. A disadvantage of using this topology is that maintenance can be very challenging and it will require constant supervision because of redundancy present in the network. I have achieved to learn about what routers and cables are used in a full mesh topology, the advantages and disadvantages of using this topology and how to configure the application and profile definitions.

Figure 16 – Full Mesh topology

4.6 Partial Mesh topology

I built this partial mesh topology with 6 Juniper T640 routers and the connection used to connect all of them was 100baseT. Also I added the application and profile definition because these are vital in a network. In a partial mesh topology, some routers are prearranged in a full mesh format but others are only connected to one or two in the network. This topology uses a small number of connections and though less expensive it is also less fault-tolerant. An advantage of using a partial mesh topology is that fewer cables are used therefore saving cost and easier to manage. A disadvantage of using this topology is that not every router is connected to each other so it will make it harder to pass data and this could be time consuming. I have achieved to learn about what routers and cables are used in a partial mesh topology, the advantages and disadvantages of using this topology and how to configure the application and profile definitions.

Figure 17 – Partial Mesh topology

4.7 Line topology

I built this line topology with 6 Juniper T640 routers and the connection used to connect all of them was 100baseT. Also I added the application and profile definition because these are vital in a network. In this topology all of the routers are connected to the main cable. So if the data is being sent between the routers then other routers cannot transmit. Also if too many routers are connected then the time taken for the data to transfer slows down very quickly as the routers have to wait longer for the line to be clear. Some of the advantages of using this topology are that it is easy to implement and extend, also it is designed for temporary networks and primarily is cheaper than other topologies. There are quite a few disadvantages of using this topology and they are although this topology can handle large networks, this can be a warning because all of the routers are connected to the same main cable, therefore increasing the number of routers in the network will have a big impact on the networks speed and quality. Also if more routers are added onto the network, more data will be required to be transmitted and because all of the routers are connected with the same main cable, all of the data has to travel down this single cable. I have achieved to learn about what routers and cables are used in a line topology, the advantages and disadvantages of using this topology and how to configure the application and profile definitions.

Figure 18 – Line topology

5. Implementation

In this part I am going to show screenshots of the configuration and what values were changed and how to configure and run the event simulation.

This screenshot shows the application configuration I created an application called video and changed the video conferencing tab to high resolution video.

Figure 19 – Application configuration

This screenshot shows the profile configuration which I set to VI and changed the application to video. These user profiles can then be specified on different nodes in the network to generate application layer traffic.

Figure 20 – Profile configuration

This screenshot shows the failure and recovery, I added 4 rows in which I ticked router 6 to router 1. Also I set the failure time to 240 seconds and 620 seconds, whereas for the recovery time it was set to 500 seconds and 800 seconds. These figures were consistent throughout my testing.

Figure 21 – Failure recovery configuration

The three protocols OSPF, RIP and EIGRP are set with its IP dynamic routing protocol. Also I changed the IP routing table to export and IP version preference to IPv4.

Figure 22 – Global attributes IP routing

I enabled the simulation efficiency for each protocol because this would make sure the results run until the end of the 20 minutes. Also the stop time is set for OSPF, RIP and EIGRP to 1200, 260 and 1000 seconds.

Figure 23 – Simulation stop for OSPF, RIP and EIGRP and run simulation for 20 minutes

For the comparison to be done I had to set the individual statistics. I want to view the results of OSPF, RIP and EIGRP for each topology. It shows the comparison of Convergence Activity and traffic sent (bits/sec). The following three figures show the statistics for showing the results.

Figure 24 – OSPF DES statistics

Figure 25 – RIP DES statistics

Figure 26 – EIGRP DES statistics

This screenshot shows the default parameters for OSPF routers.

Figure 27 – OSPF parameters

This screenshot shows the default parameters for RIP routers.

Figure 28 – RIP parameters

6. Testing and Results

6.1 Comparison 1

So when I ran the simulation the results were presented in a form of a graph. The first comparison that I am going to do is with the star topology. I want to see which protocol is better OSPF, RIP or EIGRP. The results for OSPF are in red, RIP are in green and EIGRP are in blue. From left of the graph to the right side it shows the initial peak of the traffic sent in 20 minutes. The graph shows that OSPF has the highest traffic, this is because OSPF collects a lot of data. Also I can see from the results that EIGRP has the highest bandwidth efficiency compared to RIP and OSPF.

Figure 29 – Overlaid Traffic Sent (bits/sec) on Star Topology

So when I ran the simulation the results were presented in a form of a graph. The first comparison that I am going to do is with the star topology. I want to see which protocol is better OSPF, RIP or EIGRP. The results for OSPF are in red, RIP are in green and EIGRP are in blue. From left of the graph to the right side it shows the initial peak of the convergence activity. EIGRP looks like it is quick and efficient because it reacts straight away whereas RIP is slow and OSPF is even slower.

Figure 30 – Overlaid Convergence Activity on Star Topology

6.2 Comparison 2

The figure below shows OSPF has the highest traffic, this is because OSPF uses link state and has to collect much more data than RIP and EIGRP.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 31 – Overlaid Traffic Sent (bits/sec) on Ring Topology

The screenshot below shows EIGRP is the fastest protocol against RIP and OSPF. Compared to star topology ring topology is the same, nothing has changed according to the graph.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 32 – Overlaid Convergence Activity on Ring Topology

6.3 Comparison 3

The figure below shows OSPF has the highest traffic, this is because OSPF uses the SPF algorithm to calculate the best path whereas RIP uses distance vector algorithm to calculate best path. Looking at bandwidth efficiency OSPF and EIGRP have higher efficiencies than RIP, but then RIP gets better than EIGRP.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 33 – Overlaid Traffic Sent (bits/sec) on Mesh Topology

The screenshot below shows that EIGRP is the fast protocol than RIP and OSPF, so there is no difference even though the topology is different.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 34 – Overlaid Convergence Activity on Mesh Topology

6.4 Comparison 4

The screenshot below shows that OSPF has still the highest traffic even though the topology is a line topology. EIGRP bandwidth efficiency is better than RIP and OSPF.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 35 – Overlaid Traffic Sent (bits/sec) on Line Topology

The screenshot below shows that EIGRP is still the fastest yet this is the fourth test with a different topology.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 36 – Overlaid Convergence Activity on Line Topology

6.5 Comparison 5

The screenshot below is the last comparison, so even in a partial mesh topology OSPF has the highest traffic compared to RIP and EIGRP.

 

 

 

Figure 37 – Overlaid Traffic Sent (bits/sec) on Partial Mesh Topology

The screenshot below shows EIGRP is still the fastest against RIP and OSPF, this is good because it is easy to configure and manage.

Figure 38 – Overlaid Convergence Activity on Partial Mesh Topology

7. Evaluation

7.1 Critical analysis of results

The results which I have received through the graphs are going to be evaluated in this section. So according to the traffic sent overall OSPF has the highest traffic sent compared to RIP and EIGRP. This is because OSPF uses the shortest path and it consumes less bandwidth than RIP and EIGRP. The overall findings for the convergence activity throughout the 5 topologies is that EIGRP is the fastest protocol because it has the least delay time from failing and recovering time which are set at 240 seconds and 500 seconds. To summarise EIGRP is the best protocol no matter what topology is used, but the disadvantage of using it is that there is slow convergence and limited scalability. In today’s world OSPF is used and RIP is not the best protocol to use because the performance is very slow. I have achieved my aims and I am happy but there is still more configuration in which I could do such as looking at the delay and number of hops in a topology. Lastly, I would use OSPF because it can build a hierarchical and scalable network and it supports traffic engineering. Also it has many other benefits such as it is easy to configure and manage if it was used by a large organisation.

7.2 Discussion

In general the project went well because I worked hard and hopefully I deserve a good grade for this project. But there has been some challenges while conducting this project because it is a main focus in Computer Science in today’s world. Networking in general is a challenging topic always because it requires prior knowledge as nobody can just learn everything in one go. This was good for me because I had previous knowledge from Network protocols and architectures and Computers systems security where I learned Wireshark and Linux. Also I used various tools and commands, so that has really helped me to complete this project. Otherwise you cannot simple complete this project without research and some sort of practical skills with cables and routers. In future I would compare a lot of protocols and maybe create two wide area networks and have routers, switches and pcs connected and then compare them to each other. This time round I used Juniper routers but in future I could use Cisco routers which I have outlined in my research. The best network will be to create a company with many offices around the world, but then I would like to see the network connectivity and how data reaches different offices around the world. If I do master’s degree in the future I might do something to do with WANs because this is a major topic in networking. Learning about this will put me in a position where employers look at me because I have the necessary skills such as implementing and configuring a network. This would benefit me a lot whilst looking for work in the computing industry. Finally the timeline of this project has been challenging however interesting because I have learned so much in just 10-12 weeks.

7.3 Conclusion

In conclusion, I think personally this project went well but it could’ve been better if I had put more effort and time into it. The research that I conducted was basic and I could’ve shown some calculations done through the protocols so that the research looks more sophisticated. Moving onto the literature review, there wasn’t that much previous research done on this topic so I found it hard to review some articles. The design part of this project was good because Riverbed Modeler allows you to choose any router with any link to create a network and configure using whichever protocols you prefer. Also when doing the configuration there are many parameters which can be set and the delay and time in which you want the simulation to run can be changed to your personal needs. Traffic can be monitored and there are graphs that will show the traffic being sent and received, this makes it easy to see results and evaluate from there. My testing stage went pretty well because I only chose to test the traffic sent and the network convergence activity this is because I didn’t want to overload the testing because then the graphs wouldn’t look well-ordered and it wouldn’t be clear to tell what result is what. The analysis of results are done in the evaluation part which went well also because I tested 5 topologies against three protocols, which showed good results in the graphs that were presented. If I had more time I would’ve made graphs which work out the average of the traffic sent and received of the protocols and then I would’ve known which protocol is the best to use and which protocol isn’t so good. Lastly there is the opportunity to view results on their own or overlaid, I chose to use overlaid because then you can see the comparison clearly and also the protocols are in different colours which make the results stand out.

7.4 Successes and failures

Whatever you do in life you will always have success and failure, but from failure I learnt in this project and the next time I will learn from my mistakes and make sure I produce a good project in the future. The first success I had was that there was a lot of books and articles which I looked through which made my knowledge expand on networking. The second success which I had was that I could’ve designed which ever network I want, for example a LAN, WAN or VLAN with the network simulator that I was working with. This is good because I could’ve compared these with the topologies and could’ve got some good results. The third success which I am happy with is the configuration because it is easier than command line, in Riverbed Modeler you can just click button and edit parameters, rather than in Cisco Packet Tracer you have to use command line which could make it harder to work as any error could be made easily. I had some failures with this project and the first one is that my initial project was on MPLS and RSVP but they didn’t seem to work so I had no choice to move on and choose something in which works and produces results so that I can then analyse the results and see which protocol is the best. Another aspect on this project in which I failed was to follow the Riverbed Modeler guide in which it shows you how to do everything in the network simulator. I found this after I had done my testing so it was good looking through it. But in future I will make use of books and articles before conducting with my project because research is the main part of a project because that enhances your knowledge and makes it easier to complete your project.

7.5 Management of the project

According to my understanding I am going to tell the truth when it was January 2017 I had many ideas for a project but I didn’t have the necessary skills, for example I wanted to create a racing game but then again Java programming is my strongest skill. So I was at a big disadvantage from not having enough knowledge on Java programming. I’ve always wanted to create a game which is practical and much more fun than creating networks and with Java there will be a lot of debugging to do and editing code so that the game works successfully. Back in January I had created a Gantt chart for this project but I didn’t meet the deadlines that were planned so that wasn’t helpful to me. I struggled in the beginning and then when I looked into the topic which my supervisor advised me on which was the protocols it didn’t seem hard to implement, but it was just time consuming to design and implement a network and a lot of research was required to make sure I understood what I was doing and why I am doing so. So 4 weeks ago the way I managed my project was that I created a timetable on my bedroom wall so I had wrote 5 hour slots during the day on my timetable so that I know when I am working on my project and how much I need to complete within a certain amount of time. Also this project was managed well because most of it was done on my laptop that is because I installed Riverbed Modeler Academic Edition 17.5 with the help of my supervisor. If I hadn’t then I would not be able to proceed with my project as the university didn’t have the software installed due to license issues, but then after a while it was installed in the computer science labs.

7.6 Future improvements

There are many improvements that I would do in the future, if I had another 6 months this project would be researched in depth and analysed accordingly. Also I would learn more about Riverbed Modeler so that I have a broad understanding of how to use the tools and features more efficiently. Where I have created topologies, in future I will create a LAN and a WAN and compare these using MPLS and RSVP as I didn’t have the correct version of Riverbed Modeler at the moment so it wasn’t compatible to use these protocols. Another improvement with this project that I would do is maybe go on a course like CCNA or Juniper networks and increase my knowledge on networking so that I have the necessary skills required in order to complete a fantastic project in networking. Another future improvement that I would like to complete is a Masters in networking because this is vital for when applying for jobs and I would then be able to research more on this topic and write another report on other protocols and monitor the traffic sent and received through them. Also if I had more time I would of used Cisco Packet Tracer as well to monitor activity in different topologies and the traffic sent, this would be a good idea because most of the configuration is done through command line which I have precious experience whilst studying Network protocols and architectures module. Lastly, the final improvement that I would do to this project is rewrite the report and do more background research on network protocols, maybe read some books and analyse previous work done by someone else so that I have an idea to which extent I have to work towards. Also time is a major part in a project because the more time the more research can be carried out along with more experiments to successfully solve a problem.

8. References

Aboelela, E. and Peterson, L., 2012. Network simulation experiments manual. 1st ed. Amsterdam: Morgan Kaufmann.

Alvarez, S., 2017. Chapter 4: Cisco MPLS Traffic Engineering. [online] Network World. Available at: <http://www.networkworld.com/article/2281049/lan-wan/chapter-4–cisco-mpls-traffic-engineering.html> [Accessed 25 Apr. 2017].

Anon, 2017. [online] Available at: <http://file:///C:/Users/fezaan_20/Downloads/Introduction%20to%20Modeler.pdf> [Accessed 25 Apr. 2017].

Anon, 2017. 1st ed. [ebook] Available at: <http://www.cisco.com/c/en/us/solutions/collateral/enterprise-networks/cisco-digital-network-architecture/solution-overview-c22-736580.pdf> [Accessed 25 Apr. 2017].

Anon, 2017. [online] Available at: <http://www.cs.toronto.edu/~marbach/COURSES/CSC358_S14/classification.pdf> [Accessed 25 Apr. 2017].

Anon, 2017. [online] Available at: <https://www.ijircce.com/upload/2016/june/272_Analysing.pdf> [Accessed 25 Apr. 2017].

Anon, 2017. [online] Available at: <http://www.idc-online.com/technical_references/pdfs/data_communications/Tree_Topology_Advantages_and_Disadvantages.pdf> [Accessed 25 Apr. 2017].

Anon, 2017. Advantages and Disadvantages of Using Mesh Topology. [online] Networking-basics.net. Available at: <http://www.networking-basics.net/mesh-topology/> [Accessed 25 Apr. 2017].

Anon, 2017. Auto-bandwidth for RSVP LSPs. [online] Brocade.com. Available at: <http://www.brocade.com/content/html/en/configuration-guide/netiron-05900-mplsguide/GUID-B50573E8-0079-4042-9207-2FC8F6104820.html> [Accessed 25 Apr. 2017].

Anon, 2017. CCNA Training » WAN Tutorial. [online] 9tut.com. Available at: <http://www.9tut.com/wan-tutorial> [Accessed 25 Apr. 2017].

Anon, 2017. Cite a Website – Cite This For Me. [online] Cisco.com. Available at: <http://www.cisco.com/c/dam/en/us/support/docs/ip/enhanced-interior-gateway-routing-protocol-eigrp/16406-eigrp-toc-09.gif> [Accessed 25 Apr. 2017].

Anon, 2017. configure mpls riverbed modeler – YouTube. [online] Youtube.com. Available at: <https://www.youtube.com/results?search_query=configure+mpls+riverbed+modeler&spfreload=1> [Accessed 25 Apr. 2017].

Anon, 2017. Dissertation Methodology | Advice | Postgrad.com. [online] Postgrad.com. Available at: <https://www.postgrad.com/advice/exams/dissertations_and_theses/dissertation_methodology/> [Accessed 25 Apr. 2017].

Anon, 2017. Fully Connected Network Topology Diagram | Network Topologies | Hybrid Network Topology | Network Topologies. [online] http://www.conceptdraw.com. Available at: <http://www.conceptdraw.com/examples/network-topologies> [Accessed 25 Apr. 2017].

Anon, 2017. Introduction to Information Technology – CSE1010e. [online] Vcampus.uom.ac.mu. Available at: <http://vcampus.uom.ac.mu/cse1010e/chapter_9/c9_Section12_3.htm> [Accessed 25 Apr. 2017].

Anon, 2017. Juniper T-Series. [online] En.wikipedia.org. Available at: <https://en.wikipedia.org/wiki/Juniper_T-Series> [Accessed 25 Apr. 2017].

Anon, 2017. Network Topologies – Star, Bus (Line) and Ring. [online] Klbict.co.uk. Available at: <http://www.klbict.co.uk/gcse/theory/5_3/5_3_4_topologies.htm> [Accessed 25 Apr. 2017].

Anon, 2017. OpenAthens / Sign in. [online] Ebookcentral.proquest.com. Available at: <http://ebookcentral.proquest.com/lib/herts/detail.action?docID=534935> [Accessed 25 Apr. 2017].

Anon, 2017. OpenAthens / Sign in. [online] Ebookcentral.proquest.com. Available at: <http://ebookcentral.proquest.com/lib/herts/reader.action?docID=1107581&ppg=368> [Accessed 25 Apr. 2017].

Anon, 2017. OSPF default route in JUNOS «  1. [online] Backreference.org. Available at: <http://backreference.org/2009/12/04/ospf-default-route-in-junos/> [Accessed 25 Apr. 2017].

Anon, 2017. OSPF Support for Traffic Engineering – Technical Documentation – Support – Juniper Networks. [online] Juniper.net. Available at: <https://www.juniper.net/documentation/en_US/junos/topics/concept/ospf-traffic-engineering-support-overview.html> [Accessed 25 Apr. 2017].

Anon, 2017. Routers Compare Products and Solutions. [online] Cisco. Available at: <http://www.cisco.com/c/en_uk/products/routers/buyers-guide.html#~wankeyfeatures> [Accessed 25 Apr. 2017].

Anon, 2017. TheTechnicalstuff   – How does Routing Information Protocol works. [online] Thetechnicalstuff.com. Available at: <http://www.thetechnicalstuff.com/how-does-routing-information-protocol-works/> [Accessed 25 Apr. 2017].

Anon, 2017. traffic engineering mpls – Google Search. [online] Google.co.uk. Available at: <https://www.google.co.uk/search?q=traffic+engineering+mpls&rlz=1C1IRFH_enGB607GB607&espv=2&source=lnms&tbm=isch&sa=X&sqi=2&ved=0ahUKEwjr_aDiwp_TAhWlJMAKHdaeA20Q_AUIBigB&biw=1366&bih=638#imgrc=MlHpp5Gb5kmtJM:> [Accessed 25 Apr. 2017].

Anon, 2017. Understanding the RSVP Signaling Protocol – Technical Documentation – Support – Juniper Networks. [online] Juniper.net. Available at: <http://www.juniper.net/techpubs/en_US/junos/topics/concept/mpls-security-rsvp-signaling-protocol-understanding.html> [Accessed 25 Apr. 2017].

Anon, 2017. What is WAN (wide area network)? – Definition from WhatIs.com. [online] SearchEnterpriseWAN. Available at: <http://searchenterprisewan.techtarget.com/definition/WAN> [Accessed 25 Apr. 2017].

Anon, 2017. 10 Benefits of MPLS for Large Businesses. [online] News. Available at: <http://www.elitetele.com/news/read/10-reasons-to-switch-to-mpls> [Accessed 25 Apr. 2017].

Cheang, A., 2017. Network Topology, Architecture & Protocol. [online] Aineecheang.blogspot.co.uk. Available at: <http://aineecheang.blogspot.co.uk/2012/04/network-topology-architecture-protocol.html> [Accessed 25 Apr. 2017].

Giri, 2017. Types of Network Architecture. [online] Slideshare.net. Available at: <https://www.slideshare.net/subisabari/types-of-network-architecture> [Accessed 25 Apr. 2017].

Pana, F. and Put, F., 2014. Performance evaluation of RSVP using OPNET Modeler. Simulation Modelling Practice and Theory, 49, pp.85-97.

Peterson, L. and Davie, B., 2012. Computer networks. 1st ed. Amsterdam: Morgan Kaufmann.

Price-Evans, I., 2017. What is Routing Information Protocol (RIP)?. [online] Metaswitch.com. Available at: <http://www.metaswitch.com/resources/what-is-routing-information-protocol-rip> [Accessed 25 Apr. 2017].

Sethi, A. and Hnatyshin, V., 2013. The practical OPNET user guide for computer network simulation. 1st ed. Boca Raton, Fla.: CRC Press.

Support, P., Products, E., (End-of-Sale), C. and Guides, C., 2017. Cisco ASR 9000 Series Aggregation Services Router MPLS Configuration Guide, Release 4.3.x  – Implementing RSVP for MPLS-TE   [Cisco IOS XR Software Release 4.3]. [online] Cisco. Available at: <http://www.cisco.com/c/en/us/td/docs/routers/asr9000/software/asr9k_r4-3/mpls/configuration/guide/b_mpls_cg43xasr9k/b_mpls_cg43asr9k_chapter_010.html> [Accessed 25 Apr. 2017].

Support, T., (MPLS), M. and TechNotes, C., 2017. Configuring MPLS Basic Traffic Engineering Using IS-IS. [online] Cisco. Available at: <http://www.cisco.com/c/en/us/support/docs/multiprotocol-label-switching-mpls/mpls/13737-mplsteisis.html#diag> [Accessed 25 Apr. 2017].

Team, B., 2017. Wired Network Topologies – Network Hardware & Assembly: LAN 102. [online] Tom’s IT Pro. Available at: <http://www.tomsitpro.com/articles/local_area_network-gigabit_ethernet-networking-nics-scott_mueller,2-263-4.html> [Accessed 25 Apr. 2017].

Tutorialology, W., 2017. Bus Topology. [online] Tutorialology. Available at: <http://www.tutorialology.com/computer-networking/bus-topology/> [Accessed 25 Apr. 2017].