DESIGN AND IMPLEMENTATION OF TACACS+ ON A SCALABLE ENTERPRISE BASED IP NETWORK A Project Presented to the faculty of the Department of Computer Science California State University, Sacramento Submitted in partial satisfaction of the requirements for the degree of MASTER OF SCIENCE in Computer Science by Aejazuddin Farooqui FALL 2012 DESIGN AND IMPLEMENTATION OF TACACS+ ON A SCALABLE ENTERPRISE BASED IP NETWORK A Project by Aejazuddin Farooqui Approved by: __________________________________, Committee Chair Dr. Jinsong Ouyang __________________________________, Second Reader Dr. Chung-E Wang ____________________________ Date ii Student: Aejazuddin Farooqui I certify that this student has met the requirements for format contained in the University format manual, and that this project is suitable for shelving in the Library and credit is to be awarded for the project. __________________________, Graduate Coordinator Dr. Nichrouz Faroughi Department of Computer Science iii ___________________ Date Abstract of DESIGN AND IMPLEMENTATION OF TACACS+ ON A SCALABLE ENTERPRISE BASED IP NETWORK by Aejazuddin Farooqui As corporate offices are geographically distributed, securing and managing networks has always been increasingly challenging. These offices have both highly confidential mission and business critical data forwarding between different sites and requires high degree of network security from all possible aspects. As more sites converge, careful design and planning must occur to assure that the quality, reliability and security of network is not affected. This project proposes a scalable TACACS+ architecture that can be implemented by Universities, Enterprises and Internet Service Providers. The proposed architecture facilitates easy integration of network domains and centralized manageability to provide Authentication, Authorization and Accounting (AAA) services to establish network connectivity to multivendor network elements. It also resolves two identified major issues. One, long end-to-end round trip network user authentication delay and second, lack of centralized manageability leading to unauthorized user-access regardless of job function. iv These two issues are addressed in two project phases. The first phase involves creating a prototype of an enterprise based end-to-end IP network. This prototype is designed and deployed with two sites represented as regional sites in different Autonomous Systems (AS). The second phase involves deployment of a centralized TACACS+ environment. The proposed project is implemented using Cisco networking appliance. With the successful implementation of these phases, a secure and centralized TACACS+ model was deployed with one single point to provide global administration. To compare network performance analysis during different time periods, a pre-production ACS server is deployed in a regional GAR site in Penang to provide AAA services. The average performance during peak time is improved by 87.1% and the average performance during off peak time is improved by 89.2%. Secondly, the average hop count for devices in Penang, where the AAA server is actually integrated, is improved by 75%. And the average hop count for devices in other sub-regions is improved by ~50%. Finally, with a hierarchal IP network design the end-to-end back and forth authentication request and response messages are specific to dedicated slaves in dedicated regional domains, hence reducing immense noise on network core layer. _______________________, Committee Chair Dr. Jinsong Ouyang _______________________ Date v ACKNOWLEDGEMENTS I would like to thank my supervisors Prof. Dr. Jinsong Ouyang and Prof. Dr. Martin Nicholes, for allowing me to work on this project in the office laboratory. I would like to thank my second reader Prof. Dr. Chung-E Wang for willing to review my project. I would like to thank my colleagues Jeff Lee and Brent Anderson for providing the resources I needed in the lab to accomplish this project. More importantly, I am indebted by the encouragement and inspiration that Prof. Ouyang and Prof Nicholes have provided to me. I learnt a lot of theoretical and practical concepts during this project. It not only enhanced my knowledge in network security, but also in terms of project scoping, project execution, demonstration of results, documentation, and successful project fulfillment. Once again, I would like to thank Prof. Ouyang and Prof. Nicholes for all the help and motivation and would like to share with them the credit of fulfillment of this project. I would also like to thank Prof. Dr. Cui Zhang for providing valuable information through Research Methodology on preparing for Master Project and Thesis especially how to recognize plagiarism. Last, but not the least, I would like to thanks my wife Uzma for providing me all support since the beginning of my master’s course. My best regards to all the people mentioned above. Further information on topics covered in this project can be obtained from Cisco Systems documentation website http://www.cisco.com/univercd. vi TABLE OF CONTENTS Page Acknowledgements ............................................................................................................ vi List of Tables ...................................................................................................................... x List of Figures ................................................................................................................... xii Chapter 1. INTRODUCTION .......................................................................................................... 1 1.1 Access Control System ....................................................................................... 2 1.2 What Is TACACS+? ........................................................................................... 3 1.2.1 Authentication ................................................................................................. 3 1.2.2 Authorization .................................................................................................. 4 1.2.3 Accounting ...................................................................................................... 4 1.3 Scope Of The Project .......................................................................................... 5 2. PROTOTYPE OF AN ENTERPRISE BASED IP NETWORK.................................... 6 2.1 Hierarchical Network Design ............................................................................. 6 2.2 IP Addressing ...................................................................................................... 7 2.3 Hardware & IOS ............................................................................................... 10 2.4 Implementation Of Distribution Layer At Site1 ............................................... 10 2.5 Implementation Of Access Layer At Site1 ....................................................... 16 2.6 Implementation Of Core Layer At Site1 ........................................................... 23 2.7 Configuring EIGRP .......................................................................................... 31 vii 2.8 Implementation Of Core Layer at Site2 ............................................................ 42 2.9 Integration Of Site1 and Site2 Using WAN Links ........................................... 43 3. DEPLOYMENT OF CENTRALIZED TACACS+ ENVIRONMENT ....................... 55 3.1 Installing Master And Slave TACACS+ Servers ............................................. 58 3.2 Building Master Server And Replicating To Slave Server ............................... 59 3.2.1 Creating Network Device Groups ................................................................. 60 3.2.2 Creating User Groups ................................................................................... 62 3.2.3 Creating Users ............................................................................................... 62 3.2.4 Replicating Database .................................................................................... 63 3.2.5 Master Setup ................................................................................................. 63 3.2.6 Slave Setup.................................................................................................... 64 3.2.7 Configuring Site1 and Site2 Network Devices With TACACS+ ................. 66 3.2.8 Testing For PASSED Authorization To Network Devices........................... 69 3.2.9 Testing For FAILED Authorization To Network Devices ........................... 73 3.2.10 Round Trip Delay For Site1 Primary TACACS+ Server ............................. 75 3.2.11 Round Trip Delay For Site1 Secondary TACACS+ Server ......................... 76 4. PRE-PRODUCTION DEPLOYMENT AND NETWORK PERFORMANCE ANALYSIS .................................................................................................................. 77 4.1 Analysis Of Round-Trip-Delay At Different Time Periods For A AAA Server Outside Region: ................................................................................................ 78 4.1.1 Analysis Of RTD During Morning Hours .................................................... 79 viii 4.1.2 Analysis Of RTD During Peak Hours .......................................................... 80 4.1.3 Analysis Of RTD During Afternoon Hours .................................................. 82 4.1.4 Analysis Of RTD During Off Hours ............................................................. 83 4.1.5 Analysis Of Average RTD During Different Time Periods ......................... 84 4.2 Analysis Of Hop Count At Different Time Periods For A AAA Server Outside Region: ............................................................................................................. 85 4.3 Deploying Regional Pre-Production AAA Servers .......................................... 86 4.4 Analysis Of Round-Trip-Delay At Different Time Periods For A AAA Server Within A Region: ............................................................................................. 87 4.4.1 Analysis Of RTD During Morning Hours .................................................... 87 4.4.2 Analysis Of RTD During Peak Hours .......................................................... 88 4.4.3 Analysis Of RTD During Afternoon Hours .................................................. 90 4.4.4 Analysis Of RTD During Off-Peak Hours.................................................... 91 4.4.5 Analysis Of Average RTD During Different Time Periods ......................... 92 4.5 Analysis Of Hop Count At Different Time Periods For A AAA Server In The Same Region: ................................................................................................... 94 4.6 Improvement In Network Performance ............................................................ 95 5. CONCLUSION ............................................................................................................ 99 Acronyms ........................................................................................................................ 100 References ....................................................................................................................... 101 ix LIST OF TABLES Tables Page 2-1 Subnetted Block Of 172.16.0.0/12 ....................................................................... 8 2-2 Subnetted Block Of 192.168.0.0/16 ..................................................................... 9 2-3 Hardware And IOS/Software Version ............................................................... 10 2-4 Management Connectivity At Distribution Layer ............................................. 11 2-5 Configuring Management VLAN On NMS ...................................................... 12 2-6 Configuring Access Layer Network Element .................................................... 18 4-1 Average Morning Hour Stats For A AAA Server Outside A Region ................ 79 4-2 Average Peak Hour Stats For A AAA Server Outside A Region ...................... 81 4-3 Average Afternoon Hour Stats For A AAA Server Outside A Region ............. 82 4-4 Average Off Peak Stats For A AAA Server Outside A Region ........................ 83 4-5 Average Stats For A AAA Server Outside A Region ........................................ 84 4-6 Average Hop Count For A AAA Server Outside A Region .............................. 85 4-7 Average Morning Hour Stats For A AAA Server Within A Region ................. 87 4-8 Average Peak Hour Stats For A AAA Server Within A Region ....................... 89 4-9 Average Afternoon Hour Stats For A AAA Server Within A Region .............. 90 4-10 Average Off Peak Stats For A AAA Server Within A Region .......................... 91 4-11 Average Stats For a AAA Server Within A Region .......................................... 92 4-12 Average Hop Count For A AAA Server Within A Region ............................... 94 4-13 Average Performance Improvement in RTD ..................................................... 95 x 4-14 Average Hop Count Improvement ..................................................................... 97 xi LIST OF FIGURES Figures Page 1-1 A Simple AAA Scenario ...................................................................................... 2 2-1 Hierarchical Network Design .............................................................................. 6 2-2 LAB Prototype integrated with WAN Connectivity between Two Sites ............ 7 2-3 Management VLAN Connectivity ..................................................................... 11 2-4 Layer1 - Connectivity ........................................................................................ 17 2-5 Segregation of Management and Client VLANs ............................................... 18 2-6 Integration of Distribution and Core Layer ....................................................... 24 2-7 Core VLAN’s ..................................................................................................... 25 2-8 Discovering Routes ............................................................................................ 33 2-9 Site2 Core VLAN .............................................................................................. 42 2-10 WAN Connectivity between Site1 and Site2 ..................................................... 44 2-11 Redistribution of EIGRP into BGP and BGP into EIGRP at Site1 ................... 51 3-1 Server VLAN 30 at Site1 ................................................................................... 56 3-2 ACS Web Interface ............................................................................................ 59 3-3 Centralized Administration and One Way Replication ..................................... 60 3-4 NDG of AAA Servers ........................................................................................ 61 3-5 Lab NDGs with LAN &WAN Subnets ............................................................. 61 3-6 Site1 IP Range.................................................................................................... 61 3-7 User Groups ....................................................................................................... 62 xii 3-8 NAS Restrictions ............................................................................................... 62 3-9 Group Level Settings ......................................................................................... 63 3-10 Replicating Components from Master Server.................................................... 63 3-11 Replication Servers from Master ....................................................................... 64 3-12 Replicating Components Receiving from Master Server .................................. 64 3-13 Inbound Replication Master Server ................................................................... 64 3-14 Outbound Database Replication from Master Server ........................................ 65 3-15 Inbound Database Replication on Slave Server ................................................. 66 3-16 Communication between Client and TACACS+ Server ................................... 68 3-17 Authentication Logs of User AJ ........................................................................ 73 3-18 Failed Authentication Logs of User AJ ............................................................. 75 4-1 Average RTD During Morning Hours ............................................................... 80 4-2 Average RTD During Peak Hours ..................................................................... 81 4-3 Average RTD During Afternoon Hours ............................................................ 82 4-4 Average RTD During Off Hours ....................................................................... 83 4-5 Average RTD At Different Time Periods .......................................................... 84 4-6 Periodic Representation Of Hop Count From Different Regions ...................... 85 4-7 Average RTD During Morning Hours ............................................................... 88 4-8 Average RTD During Peak Hours ..................................................................... 89 4-9 Average RTD During Afternoon Hours ............................................................ 91 4-10 Average RTD During Off-Peak Hours .............................................................. 92 xiii 4-11 Average RTD At Different Time Periods .......................................................... 93 4-12 Periodic Hop Count To A TACACS+ Server In the Same Region ................... 94 4-13 Average Performance Improvement in RTD ..................................................... 96 4-14 Average Hop Count Improvement ..................................................................... 97 xiv 1 CHAPTER 1. INTRODUCTION Securing and managing networks has always been increasingly challenging in corporate offices. As network grows, with mergers and acquisitions or new site deployment, it becomes more challenging to converge and secure these decentralized networks. These offices have both highly confidential mission and business critical data forwarding between different sites and requires high degree of network security from all possible aspects. One of the ways to secure a network is to limit network access services to the authorized users based on their business need. Equally as challenging is closely monitoring what services are being used and their frequency of use. Network elements such as Routers and Switches usually use Access Control Lists (ACL) to filter the source and destination IP addresses coming in and out of the configured device interfaces. These restrictions are applied and are limited to a device and not an individual! For example, if a device is configured to allow traffic from host 10.1.1.1 to access a web server say WS1, then anyone who is sitting on the host 10.1.1.1 will automatically have access to WS1. To prevent this, a more secure and flexible filtering method would be to provide access to specific authorized users. That means users with a dedicated username and password are only authorized to access the service or login to a network device. One solution would be to create a user database on every network element to restrict access to. However, this would become administrative overhead to update user profiles, and, very hard and laborious to manage in corporate offices where there are thousands of network elements. 2 What ideally required is a centralized Access Control System to administer and manage user database. 1.1 Access Control System Access Control System (ACS) is a scalable, high-performance Remote Access Dial in User Services (RADIUS) and Terminal Access Controller Access Control System (TACACS+) security server. As ACS acts as a centralized control point for managing enterprise network users, administrators and network infrastructure resources, it provides identity-based network-access control solution for intelligent information networks [1]. It supports a broad variety of multivendor network access devices also known as AAA clients, including wired and wireless LAN switches and access points, edge and core routers, Voice over IP (VoIP), Firewall, etc. Figure 1-1 [1] shows a simple AAA scenario where an ACS is functioning as an AAA server for network access devices. Figure 1-1 A Simple AAA Scenario The Network Access Device (NAD) functions as an AAA client from the ACS perspective, to direct all end-user host access requests to ACS, via the TACACS+ or RADIUS protocols. 3 The NAD serves as the network gatekeeper, and sends an access request to ACS on behalf of the user. ACS verifies the username, password and possibly other data by using its internal database or one of the configured external identity directories. ACS ultimately responds to the NAD with an access denied or an access-accept message with a set of authorization attributes. 1.2 What Is TACACS+? Terminal Access Controller Access Control System (TACACS) is a connection oriented Access Control Protocol (ACP) that provides authorization for network administrative operations on the network infrastructure itself. It provides separate control of each service: Authentication, Authorization and Accounting [1]. It is a client/server method that stores specific rights for users by associating attribute-value pair for each user. The AAA authorization daemon on a network device such as router or a switch communicates with the TACACS+ server to determine the correct authorization for different option, such as EXEC and network access. 1.2.1 Authentication Authentication provides a method for handling user identification, login and password dialog, challenge and response, messaging and encryption. Authentication identifies users prior to access to a network and network services [2]. The authentication facility provides the ability to conduct an arbitrary dialog with the user. For example, after a login and password are provided, to challenge a number of questions, like mother’s maiden name or service type, etc. In addition, the TACACS+ authentication service supports sending 4 messages to user screens. For example, a message could notify user that their password must be changed because of the corporate password aging policy of 90 days. 1.2.2 Authorization Authorization determines what a user is allowed to do. ACS can send user profile policies to AAA clients to determine which network services the user can access. Different users and groups can be authorization to give different levels of service. For example, level one user might not have the same access privileges as a level two user. You can also differentiate by levels of security, access times, and services. ACS access restrictions feature can be set to permit or deny logins based on time-of-day and day-of-week. For example, a group of users can be created for temporary accounts that can disable on specified dates. Users can also be restricted to a service or combination of services such as PPP, ARAP or EXEC. After a user selects a service, Layer 2 and Layer 3 protocols can be restricted, such as IP and IPX, and individual access lists can be applied. Access lists on a per-user or per-group basis can restrict users from reaching parts of the network where critical information is stored or prevent them from using certain services, such as File Transfer Protocol (FTP) or Simple Network Management Protocol (SNMP). 1.2.3 Accounting AAA clients use the accounting functions that the RADIUS and TACACS+ protocols provide to communicate relevant data for each user session to the AAA server for recording. ACS writes accounting records to a comma-separated value (CSV) log file or 5 ODBC database. These logs can be imported into popular database and spreadsheet applications for billing, security audits, and report generation. Network managers can use the accounting facility to track user activity for a security audit or to provide information for user billing. Accounting records include user identities, start and stop times, executed commands (such as PPP), number of packets, and number of bytes. 1.3 Scope Of The Project The project is divided into two phases. The first phase involves creating a prototype of an enterprise based end-to-end IP network. This prototype is designed and deployed with two sites represented as regional sites in different Autonomous Systems (AS) and interconnected via WAN links. The second phase involves deployment of a centralized TACACS+ environment using Cisco Secure ACS application. In this phase the TACACS+ design consists of a centralized ‘Master’ server and dedicated regional Primary and Secondary ‘Slave’ servers for providing AAA services. Chapter two and three describers in details about the implementation of the above two phases with working configuration and test cases. Finally, chapter four describes about pre-production deployment and network performance analysis of deploying regional ACS servers. 6 CHAPTER 2. PROTOTYPE OF AN ENTERPRISE BASED IP NETWORK In order to deploy and test TACACS+ architecture, a prototype of an enterprise based IP network is deployed as phase1. In this phase, two site networks are built and configured with different autonomous systems to represent them as different sites. 2.1 Hierarchical Network Design Each site is designed in a hierarchical model to enable network design in layers. And each layer can be focused on specific functions, thereby enabling the design engineer to choose the right systems and features for the layer. There are many other advantages of implementing a hierarchical design, such as easy to scale the network, easy to understand and troubleshoot the network, can maintain consistent design at different sites, etc. Core A CoreB EIGRP/ OSPF/ RIP CORE Layer L3 L3 DS1 DIST Layer DS2 VLAN 30 VLAN 30 L3 L3 VLAN 10 VLAN 10 VLAN 20 VLAN 20 TACACS (AAA) Master TACACS (AAA) Slave AS1 ACCESS Layer L2 AS2 L2 AS3 L2 Client 1 AS4 L2 Client 2 Figure 2-1 Hierarchical Network Design Figure 2-1 shows an example of a hierarchical network design that was implemented as one of the project sites. The design consists of a backbone (core) layer, which provides 7 optimal transport between sites. The distribution layer, which provides policy based connectivity, route summarization or aggregation and the access-layer, which provides work group or user access to the network [3]. To maintain consistency Site2 was designed similar to Site1 in hierarchical layers. And to simplify the end-to-end architecture and connectivity, the two sites were connected via point-to-point WAN links. Figure2-2 shows the point-to-point connectivity between Core routers of Site1 and Site2. Figure 2-2 LAB Prototype integrated with WAN Connectivity between Two Sites 2.2 IP Addressing Effective IP address management plays a very important role in designing large scale networks especially when networks are geographically distributed. Use of effective IP 8 addressing helps to perform route summarization at distribution layer there by reducing traffic on core, secondly easy management by allocating dedicated address space to dedicated site, and subnetting within each site. Two private address spaces 172.16.0.0/12 and 192.168.0.0/16, as specified in RFC 1918, were used to implement a hierarchical design and keep the internal and external routes distinctive. The address block 172.16.0.0/12 is used to implement WAN and the address block 192.168.0.0/16 is used to implement LAN design. The tables 2-1 shows the two major blocks represented in bit format. Table 2-1 Subnetted Block Of 172.16.0.0/12 Major Block 172.16.0.0/12 172.16.0.0 172.16.0.1 … 172.16.0.255 172.16.1.0 172.16.1.0 172.16.1.4 172.16.1.8 172.16.1.12 172.16.1.16 … … 172.16.1.255 … … 172.16.255.255 172.17.0.0 … Major Block 172.16.0.0/12 … … 172.31.255.255 Binary Representation of 2nd, 3rd and 4th Octet 172.0001 0000.0000 0000.0000 0000 172.0001 0000.0000 0000.0000 0001 172.0001 0000.0000 0000.1111 1111 172.0001 0000.0000 0001.0000 0000 172.0001 0000.0000 0001.0000 0000 172.0001 0000.0000 0001.0000 0100 172.0001 0000.0000 0001.0000 1000 172.0001 0000.0000 0001.0000 1100 172.0001 0000.0000 0001.0001 0000 Subnet Mask 255.255.255.0 255.255.255.252 255.255.255.252 255.255.255.252 255.255.255.252 255.255.255.252 172.0001 0000.0000 0001.1111 1111 172.0001 0000.1111 1111.1111 1111 172.0001 0001.0000 0000.0000 0000 Binary Representation of 2nd, 3rd and 4th Octet 172.0001 1111.1111 1111.1111 1111 Subnet Mask 9 Table 2-2 Subnetted Block Of 192.168.0.0/16 Major Block 192.168.0.0/16 192.168.0.0 192.168.1.0/24 192.168.1.1 192.168.1.2 … … 192.168.1.255 Binary Representation of 2nd, 3rd and 4th Octet 192.1010 1000.0000 0000.0000 0000 192.1010 1000.0000 0001.0000 0000 192.1010 1000.0000 0001.0000 0001 192.1010 1000.0000 0001.0000 0010 192.168.2.0/24 192.168.3.0/24 192.168.4.0/24 192.168.5.0/24 … … … 192.168.250.0/24 192.168.250.0/27 192.168.250.32/2 7 192.168.250.64/2 7 192.168.250.96/2 7 192.168.250.128/ 27 … … 192.168.250.224/ 27 192.168.251.0/24 … 192.168.255.0/24 192.168.255.1 … 192.168.255.255 192.1010 1000.0000 0010.0000 0000 192.1010 1000.0000 0011.0000 0000 192.1010 1000.0000 0100.0000 0000 192.1010 1000.0000 0101.0000 0000 255.255.255.0 255.255.255.0 255.255.255.0 255.255.255.0 192.1010 1000.1111 1010.0000 0000 192.1010 1000.1111 1010.0000 0000 192.1010 1000.1111 1010.0010 0000 255.255.255.0 255.255.255.224 255.255.255.224 192.1010 1000.1111 1010.0100 0000 255.255.255.224 192.1010 1000.1111 1010.0110 0000 255.255.255.224 192.1010 1000.1111 1010.1000 0000 255.255.255.224 192.1010 1000.1111 1010.1110 0000 255.255.255.224 192.1010 1000.1111 1011.0000 0000 255.255.255.0 192.1010 1000.1111 1111.0000 0000 192.1010 1000.1111 1111.0000 0001 255.255.255.0 Subnet Mask 255.255.255.0 192.1010 1000.0000 0001.1111 1111 192.1010 1000.1111 1111.1111 1111 10 The table2-2 shows the subnetted block of 192.168.0.0/16 which was reserved for LAN development. Several /24 subnets were reserved for Client and Management VLAN’s. A dedicated 192.168.250.0/24 subnet was further subnetted to /27 subnets to establish EIGRP connectivity between the distribution layer and the core layer among different sites. As a /27 subnet can host up to 30 IP’s it can scale the site distribution layer to connect 30 distribution layer 3 switches or routers. 2.3 Hardware & IOS Several devices and IOS’s were used at different layers of the hierarchical design and the Cisco 3640 routers were upgraded from 11.2 IOS version to 12.2 IOS version to allow advance CLI features. The following table 2-3 shows the hardware and IOS version used to implement the prototype. Table 2-3 Hardware And IOS/Software Version Hardware Catalyst 6500 Sup 720 Cisco 3640 Cisco 2950 HP G3 Server 2.4 IOS Version 12.2(33)SXH5 12.2(27) 12.1(22) Win 2003 Server Cisco Secure ACS 4.2 Implementation Of Distribution Layer At Site1 The distribution layer is the demarcation point between the access and core layer and help to define and differentiate the core. In a site or campus design, the distribution layer can include several functions such as address or area aggregation, workgroup access, broadcast/multicast domain definitions, VLAN routing, and security [3]. 11 A pair of redundant, Catalyst 6509 Layer3, switches are integrated as distribution layer switches. As shown in Figure2-3, these switches are configured with an IP address from a dedicated management VLAN 50 to provide management connectivity. Figure 2-3 Management VLAN Connectivity A dedicated Network Management Switch (NMS) is configured to provide separate management connectivity for all devices in a campus LAN. The links are configured as access links and no other VLAN is allowed except the management VLAN 50. The Tables 2-4 and 2-5 shows the configuration of the Distribution and the NMS switches. A subnet 192.168.1.0/24 is reserved to provide management connectivity. Table 2-4 Management Connectivity At Distribution Layer Distribution Switch 1 ! vlan 50 name AJ_Lab_Mgmt_Vlan ! Distribution Switch2 ! vlan 50 name AJ_Lab_Mgmt_Vlan ! 12 interface Vlan50 description AJ_Lab_Mgmt_Vlan ip address 192.168.1.3 255.255.255.0 no ip redirects standby 50 ip 192.168.1.1 standby 50 timers 1 3 standby 50 priority 10 standby 50 preempt ! interface GigabitEthernet2/2 description NMS_Port0/1 no ip address switchport switchport access vlan 50 switchport mode access end interface Vlan50 description AJ_Lab_Mgmt_Vlan ip address 192.168.1.2 255.255.255.0 no ip redirects standby 50 ip 192.168.1.1 standby 50 timers 1 3 standby 50 priority 15 standby 50 preempt ! interface GigabitEthernet7/2 description NMS_Port0/2 no ip address switchport switchport access vlan 50 switchport mode access end Table 2-5 Configuring Management VLAN On NMS Network Management Switch ! vlan 50 name AJ_Lab_Mgmt_Vlan ! interface Vlan50 ip address 192.168.1.5 255.255.255.0 no ip route-cache ! interface GigabitEthernet0/1 switchport access vlan 50 switchport mode access ! interface GigabitEthernet0/2 switchport access vlan 50 switchport mode access end Test: Connectivity Test from NMS to DS1 and DS2 Result: 100%Ping Success Rate to the interfaces on DS1 and DS2. NMS#ping 192.168.1.2 Type escape sequence to abort. 13 Sending 5, 100-byte ICMP Echos to 192.168.1.2, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/203/1012 ms NMS#ping 192.168.1.3 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.1.3, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/200/1000 ms NMS# To provide 100% network uptime for any given VLAN, HSRP is implementing and a Virtual IP 192.168.1.1 is configured as a sharing IP between the two distribution switches. For each VLAN, distribution switch is either Active or Standby depending on the priority. A priority value of 25 is configured as Active and a priority value of 10 is configured as Standby. A group ID 50 is assigned to VLAN50 and the members of this virtual group continuously exchange status messages. This way one device assumes the routing responsibility of other, should it go out of communication. An ether-channel, P01, as show in Figure 2-3 is configured between the two distribution pair to serve as either a trunk to exchange VTP information or to serve as a high availability link. The following configuration is applied to configure the Ether-Channel and HSRP on distribution layer switches: Distribution Switch 1 ! interface Vlan50 standby 50 ip 192.168.1.1 standby 50 timers 1 3 standby 50 priority 10 standby 50 preempt ! ! interface Port-channel1 switchport Distribution Switch2 ! interface Vlan50 standby 50 ip 192.168.1.1 standby 50 timers 1 3 standby 50 priority 25 standby 50 preempt ! ! interface Port-channel1 switchport 14 switchport trunk encapsulation dot1q switchport trunk native vlan 999 switchport trunk allowed vlan 2-9,1149,51-99,101-199,201-998,1000-4094 switchport mode trunk storm-control broadcast level 0.50 end ! interface GigabitEthernet2/3 description Back-To-Back Link DS2_6/3 channel-group 1 mode desirable ! switchport trunk encapsulation dot1q switchport trunk native vlan 999 switchport trunk allowed vlan 2-9,1149,51-99,101-199,201-998,1000-4094 switchport mode trunk storm-control broadcast level 0.50 end ! interface GigabitEthernet6/3 description Back-To-Back Link DS1_2/3 channel-group 1 mode desirable ! Test: Ether-Channel Check on DS1 and DS2 Result: The output shows an Ether-Channel also called as Port-Channel P01 is created with a Group ID 1 and the port 2/3 in DS1 and port 6/3 in DS2 are in port-channel 1. DS1#sh etherchannel summary Flags: D - down P - in port-channel I - stand-alone s - suspended H - Hot-standby (LACP only) R - Layer3 S - Layer2 U - in use f - failed to allocate aggregator u - unsuitable for bundling Number of channel-groups in use: 1 Number of aggregators: 1 Group Port-channel Protocol Ports ------+-------------+-----------+----------------------------------------------1 Po1(SU) PAgP Gi2/3(P) DS1# DS2#sh etherchannel summary Flags: D - down P - bundled in port-channel I - stand-alone s - suspended H - Hot-standby (LACP only) R - Layer3 S - Layer2 U - in use N - not in use, no aggregation f - failed to allocate aggregator M - not in use, no aggregation due to minimum links not met 15 m - not in use, port not aggregated due to minimum links not met u - unsuitable for bundling d - default port w - waiting to be aggregated Number of channel-groups in use: 1 Number of aggregators: 1 Group Port-channel Protocol Ports ------+-------------+-----------+----------------------------------------------1 Po1(SU) PAgP Gi6/3(P) Test: Active & Standby test for HSRP at DS1 and DS2 Result: 100% ping result from NMS to Virtual IP 192.168.1.1 configured on DS1 and DS2. Secondly, DS1 should be standby and DS2 as Active for VLAN 50. NMS#ping 192.168.1.1 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.1.1, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/4/12 ms DS1#sh standby br P indicates configured to preempt. | Interface Grp Prio P State Active addr Standby addr Group addr Vl50 50 10 P Standby 192.168.1.2 local 192.168.1.1 DS1# DS2#sh standby br P indicates configured to preempt. | Interface Grp Prio P State Active Standby Virtual IP Vl50 50 25 P Active local 192.168.1.3 192.168.1.1 DS2# Test: HSRP Failover - Shutdown interface vlan 50 on DS2 Result: DS1 should become active and the Virtual IP should be reachable. 16 Distribution Switch 2 DS2(config)#int vlan 50 DS2(config-if)#shut DS2(config-if)# 10w2d: %HSRP-5-STATECHANGE: Vlan50 Grp 50 state Active -> Init Distribution Switch 1 10w2d: %STANDBY-6STATECHANGE: Vlan50 Group 50 state Standby -> Active As interface VLAN 50 is shut on DS2, a message a generated on DS2 and DS1 about the change of state. Observe the State change from Standby to Active on DS1. DS1#sh stan br P indicates configured to preempt. | Interface Grp Prio P State Active addr Standby addr Group addr Vl50 50 10 P Active local unknown 192.168.1.1 DS1# DS2#sh standby br P indicates configured to preempt. | Interface Grp Prio P State Active Standby Virtual IP Vl50 50 25 P Init unknown unknown 192.168.1.1 DS2# Ping the Virtual IP interface which is active on DS1 from NMS. Result: 100% Ping Success Rage. NMS#ping 192.168.1.1 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.1.1, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms NMS# 2.5 Implementation Of Access Layer At Site1 Access Layer is the layer where clients are connected or where local users are allowed into the network. A Cisco 2950 switch is installed and commissioned as an Access Layer switch. As shown in Figure 2-4, two fibre Gigabit Ethernet links one to each Distribution 17 Layer Switch are connected and configured as Trunks and a Fast Ethernet connection to NMS for Network Management Access to the switch. G6/3 G2/3 G2/2 G7/1 G2/1 G7/2 Distribution-Switch 2 6509 Distribution-Switch 1 6509 G0/1 G0/2 TRUNK TRUNK Fa0/5 G0/1 NMS 2950 G0/2 Fa0/5 NMS Connectivity AS1 2950 Figure 2-4 Layer1 - Connectivity An unused IP from the management subnet 192.168.1.0/24 is used to configure the access switch for management connectivity. A client Vlan10 is created and a subnet 192.168.3.0/24 is allocated for this Vlan. Layer 3 interfaces for Vlan10 are created on DS1 and DS2 with the first IP address 192.168.3.1 configured as Virtual IP. This virtual IP will act as default gateway to the clients. As shown in Figure 2-5, to segregates the management and the client traffic, the VLAN 10 is allowed through the uplink trunks to the distribution layer switches and the management VLAN 50 is allowed through the FE link to NMS. 18 P01 P01 G7/1 G2/2 G2/1 Distribution-Switch 1 6509 Managment VLAN50 Access VLAN10 G7/2 Distribution-Switch 2 6509 192.168.1.0/24 192.168.3.0/24 Client Traffic NMS 2950 Management Traffic Access-Switch 1 2950 Client A Client B Figure 2-5 Segregation of Management and Client VLANs The table 2-6 shows the configuration and the connectivity to the Access Layer switch. Table 2-6 Configuring Access Layer Network Element ! vlan 10 name AJ_Lab_Client_Vlan10 ! vlan 50 name AJ_Lab_Mgmt_Vlan ! vlan 999 name AJ_Lab_Native_Vlan ! interface Vlan50 description Management Connectivity 19 ip address 192.168.1.4 255.255.255.0 no ip route-cache ! interface FastEthernet0/5 description NMS _Fa0/5 switchport access vlan 50 switchport mode access no ip address ! interface GigabitEthernet0/1 description DS1_G2/1 switchport trunk native vlan 999 switchport trunk allowed vlan 10 switchport mode trunk no ip address ! interface GigabitEthernet0/2 description DS2_G7/1 switchport trunk native vlan 999 switchport trunk allowed vlan 10 switchport mode trunk no ip address ! ! interface FastEthernet0/10 description Client A switchport access vlan 10 switchport mode access no ip address ! ip default-gateway 192.168.1.1 ! Configuring the interface Fa0/5 on NMS connected to the AS1. ! interface FastEthernet0/5 switchport access vlan 50 switchport mode access ! 20 Configuring the VLAN’s and the Layer 3 interfaces on DS1 and DS2 and allowing the VLAN 10 through the trunk facing AS1. Distribution Switch 1 ! vlan 10 name AJ_Lab_Client_Vlan10 ! ! interface Vlan10 description Client VLAN ip address 192.168.3.3 255.255.255.0 no ip redirects standby 10 ip 192.168.3.1 standby 10 timers 1 3 standby 10 priority 25 standby 10 preempt ! ! interface GigabitEthernet2/1 description connected to AS1_G0/1 no ip address switchport switchport trunk encapsulation dot1q switchport trunk native vlan 999 switchport trunk allowed vlan 10 switchport mode trunk end Distribution Switch 2 ! vlan 10 name AJ_Lab_Client_Vlan10 ! ! interface Vlan10 description Client VLAN ip address 192.168.3.2 255.255.255.0 no ip redirects standby 10 ip 192.168.3.1 standby 10 timers 1 3 standby 10 priority 15 standby 10 preempt ! ! interface GigabitEthernet7/1 description connected to AS1_G0/2 no ip address switchport switchport trunk encapsulation dot1q switchport trunk native vlan 999 switchport trunk allowed vlan 10 switchport mode trunk end By the end of the above configuration the access switch should be able to ping the NMS gateway 192.168.1.1 which is a Virtual IP on VLAN 50 and it should also be able to ping the client gateway 192.168.3.1 which is a Virtual IP on VLAN 10. Test: Ensure that the physical interfaces G0/1, G0/2 and Fa0/5 on AS1 connected to Distribution Switches and the Management Switch are UP. Result: All interfaces should be UP. SW2#sh ip int br | i up Vlan50 192.168.1.4 YES NVRAM up up 21 FastEthernet0/5 GigabitEthernet0/1 GigabitEthernet0/2 SW2# unassigned YES unset up unassigned YES unset up unassigned YES unset up up up up Test: Ensure that Layer 3 interfaces for Management VLAN is reachable Result: The success rate should be 100% SW2#ping 192.168.1.1 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.1.1, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms SW2# Test: Ensure that Layer 3 interfaces for Client VLAN is reachable Result: The success rate should be 100% SW2#ping 192.168.3.1 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.3.1, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/3/8 ms SW2# Test: Ensure that DS1 is in Active State and DS2 in Standby State for VLAN 10 Result: VLAN 10 should be Active on DS1 and Standby on DS2 DS1#sh stan br P indicates configured to preempt. | Interface Grp Prio P State Active addr Standby addr Group addr Vl10 10 25 P Active local 192.168.3.2 192.168.3.1 Vl50 50 10 P Standby 192.168.1.2 local 192.168.1.1 DS1# DS2#sh stan br P indicates configured to preempt. | 22 Interface Grp Prio P State Active Standby Vl10 10 15 P Standby 192.168.3.3 local Vl50 50 25 P Active local 192.168.1.3 DS2# Virtual IP 192.168.3.1 192.168.1.1 Test: Shutdown the interface VLAN 10 on DS1 and ensure that HSRP failover works for Client VLAN 10. Result: The ping success rate to virtual IP should be 100% and DS2 should be Active HSRP gateway. DS1(config)#int vlan 10 DS1(config-if)#shut DS2# 10w2d: %HSRP-5-STATECHANGE: Vlan10 Grp 10 state Standby -> Active 10w2d: %STANDBY-6STATECHANGE: Vlan10 Group 10 state Active -> Init DS2#sh stan br P indicates configured to preempt. | Interface Grp Prio P State Active Standby Vl10 10 15 P Active local unknown Vl50 50 25 P Active local 192.168.1.3 DS2# Virtual IP 192.168.3.1 192.168.1.1 Ping result from AS1 to Virtual IP 192.168.3.1 SW2#ping 192.168.3.1 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.3.1, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/8 ms SW2# Connect a client to one of the ports on AS1. Ensure that the default gateway is reachable from the client. 23 C:\Users\aj>ipconfig Connection-specific DNS Suffix . : IPv4 Address. . . . . . . . . . . : 192.168.3.10 Subnet Mask . . . . . . . . . . . : 255.255.255.0 Default Gateway . . . . . . . . . : 192.168.3.1 C:\Users\aj>ping 192.168.3.1 Pinging 192.168.3.1 with 32 bytes of data: Reply from 192.168.3.1: bytes=32 time<1ms TTL=255 Reply from 192.168.3.1: bytes=32 time<1ms TTL=255 Reply from 192.168.3.1: bytes=32 time<1ms TTL=255 Reply from 192.168.3.1: bytes=32 time<1ms TTL=255 Ping statistics for 192.168.3.1: Packets: Sent = 4, Received = 4, Lost = 0 (0% loss), Approximate round trip times in milli-seconds: Minimum = 0ms, Maximum = 0ms, Average = 0ms C:\Users\aj> With the successful integration of access switch and also connectivity of clients to the layer 3 interface or gateway, the access layer can be extended by any number of access layer switches from the distribution layer. 2.6 Implementation Of Core Layer At Site1 The core layer is a high-speed switching backbone and should be designed to switch packets as fast as possible [3]. The figure 2-6 shows the layer1 lab core design connectivity which comprises of two Cisco 3640 routers integrated at each site. It also includes two layer 3 switches as core switches to integrate distribution layer with the core layer. Two VLAN, VLAN 100 and VLAN 200 with /27 subnets are used to provide redundant connections between distribution and core layer. Figure 2-7 shows the layer2 connectivity of LAN100 and VLAN200. A dynamic routing protocol EIGRP is used as 24 the routing protocol as it provides fast convergence with minimum network traffic and scales effectively in a well-designed network. Fa1/0 Fa1/0.1 .2 Fa1/0.2 .68 Fa1/0 Fa1/0.1 .3 Fa1/0.2 .69 Lo=192.168.5.1 .34 Fa2/0 Lo=192.168.5.2 Fa2/0 .35 Core B 3640 Core A 3640 G0/4 G0/3 G0/2 G0/1 Core-SW1 3560 EIGRP AS 10 Fa0/3 Fa0/4 G0/1 G0/2 Core-SW2 3550 Access G1/2 G1/1 .4 Lo=192.168.5.3 .36 Distribution-Switch 1 6509 G6/2 .5 G6/1 .37 Lo=192.168.5.4 Distribution-Switch 2 6509 Figure 2-6 Integration of Distribution and Core Layer Redundant fibre up links from the distribution layers switches are connected to the core switches. The core switch, Core-SW1, allow traffic for VLAN 100 with a subnet 192.168.250.0/27 and the core switch, Core-SW2, allows traffic for VLAN 200 with a subnet 192.168.250.32/27. Layer 3 interfaces for these VLAN’s are created at the core routers A and B. A subnet 192.168.5.0/24 is reserved for loopback interfaces to allow devices to participate in the EIGRP process. 25 Fa1/0 Fa1/0.1 .2 Fa1/0.2 .68 Lo=192.168.5.2 Lo=192.168.5.1 Fa2/0 Fa1/0 Fa1/0.1 .3 Fa1/0.2 .69 .34 Core A 3640 Fa2/0 .35 Core B 3640 Access VLAN200 192.168.250.32/27 Access VLAN100 192.168.250.0/27 G1/1 .4 G1/2 .36 Lo=192.168.5.3 G6/2 .5 G6/1 .37 Lo=192.168.5.4 Distribution-Switch 1 6509 Distribution-Switch 2 6509 Figure 2-7 Core VLAN’s The implementation of the core layer is initiated with the integration of the core switches. The Gigabit Ethernet links facing the distribution layer switches are configured as access links to avoid VLANs exchanging over the link and being forwarded to core. The core switches are configured as below: Core-SW1 ! vlan 100 name AJ_Lab_Core_Vlan_A ! interface GigabitEthernet0/1 description Distribution-SW1_G1/1 switchport access vlan 100 switchport mode access ! interface GigabitEthernet0/2 Core-SW2 ! vlan 200 name AJ_Lab_Core_Vlan_B ! interface GigabitEthernet0/1 description Distribution-SW1_G1/2 switchport access vlan 200 switchport mode access ! interface GigabitEthernet0/2 26 description Distribution-SW2_G6/2 switchport access vlan 100 switchport mode access ! interface GigabitEthernet0/3 description Core-A_Fa1/0.1 switchport trunk encapsulation dot1q switchport trunk allowed vlan 30,100 switchport mode trunk ! interface GigabitEthernet0/4 description Core-B_Fa1/0.1 switchport trunk encapsulation dot1q switchport trunk allowed vlan 30,100 switchport mode trunk ! description Distribution-SW2_G6/1 switchport access vlan 200 switchport mode access ! interface FastEthernet0/3 description Core-A_Fa2/0 switchport access vlan 200 switchport mode access no ip address ! interface FastEthernet0/4 description Core-B_Fa2/0 switchport access vlan 200 switchport mode access no ip address ! Now that the core switches and their interfaces are configured the next step is to configure the core routers and the layer 3 interfaces on them. The core routers are configured as: Core Router A ! interface Loopback0 ip address 192.168.5.1 255.255.255.255 ! ! interface FastEthernet1/0.1 encapsulation dot1Q 100 ip address 192.168.250.2 255.255.255.224 ! interface FastEthernet1/0.2 encapsulation dot1Q 30 ip address 192.168.250.68 255.255.255.224 ! interface FastEthernet2/0 ip address 192.168.250.34 255.255.255.224 duplex auto speed auto Core Router B ! interface Loopback0 ip address 192.168.5.2 255.255.255.255 ! ! interface FastEthernet1/0.1 encapsulation dot1Q 100 ip address 192.168.250.3 255.255.255.224 ! interface FastEthernet1/0.2 encapsulation dot1Q 30 ip address 192.168.250.69 255.255.255.224 ! interface FastEthernet2/0 ip address 192.168.250.35 255.255.255.224 duplex auto speed auto 27 ! ! Perform check to ensure that the link between the core switches and the core routers is established. Core-SW1#sh ip int br Interface IP-Address OK? Method Status Protocol Vlan1 unassigned YES unset administratively down down Vlan100 192.168.250.1 YES manual up up GigabitEthernet0/3 unassigned YES unset up up GigabitEthernet0/4 unassigned YES unset up up Core-SW2#sh ip int br Interface IP-Address OK? Method Status Protocol Vlan1 unassigned YES NVRAM administratively down down Vlan200 192.168.250.33 YES manual up up FastEthernet0/3 unassigned YES unset up up FastEthernet0/4 unassigned YES unset up up Core-R1#sh ip int br Interface IP-Address OK? Method Status Protocol FastEthernet1/0 unassigned YES NVRAM up up FastEthernet1/0.1 192.168.250.2 YES NVRAM up up FastEthernet1/0.2 192.168.250.68 YES NVRAM up up FastEthernet2/0 192.168.250.34 YES NVRAM up up FastEthernet3/0 172.16.1.1 YES NVRAM administratively down down Loopback0 192.168.5.1 YES NVRAM up up Core-R1# Core-R2#sh ip int br Interface IP-Address OK? Method Status Protocol FastEthernet1/0 unassigned YES NVRAM up up FastEthernet1/0.1 192.168.250.3 YES NVRAM up up FastEthernet1/0.2 192.168.250.69 YES NVRAM up up FastEthernet2/0 192.168.250.35 YES NVRAM up up Loopback0 192.168.5.2 YES NVRAM up up Core-R2# Ensure that the default gateway is reachable from the core switches. Core-SW1#ping 192.168.250.2 Type escape sequence to abort. 28 Sending 5, 100-byte ICMP Echos to 192.168.250.2, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/9 ms Core-SW1#ping 192.168.250.3 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.3, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/203/1006 ms Core-SW1# Core-SW2#ping 192.168.250.34 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.34, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms Core-SW2#ping 192.168.250.35 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.35, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/200/1000 ms Core-SW2# Ensure that the neighbor relationship is established between the core switches and the core routers. Core-SW1#sh cdp nei Capability Codes: R - Router, T - Trans Bridge, B - Source Route Bridge S - Switch, H - Host, I - IGMP, r - Repeater, P - Phone Device ID Local Intrfce Holdtme Capability Platform Port ID Core-R1 Gig 0/3 145 R 3640 Fas 1/0 Core-R2 Gig 0/4 121 R 3640 Fas 1/0 Core-SW1# ore-SW2#sh cdp nei Capability Codes: R - Router, T - Trans Bridge, B - Source Route Bridge S - Switch, H - Host, I - IGMP, r - Repeater, P - Phone Device ID Core-R1 Core-R2 Core-SW2# Local Intrfce Holdtme Capability Platform Port ID Fas 0/3 161 R 3640 Fas 2/0 Fas 0/4 137 R 3640 Fas 2/0 29 Configure the distribution layer3 switches to integrate with core. Redundant uplinks from each distribution pair is configured one in each core VLANs 100 and 200. Distribution – Switch 1 ! vlan 100 name AJ_Lab_Core_Vlan_A ! vlan 200 name AJ_Lab_Core_Vlan_B ! interface Vlan100 ip address 192.168.250.4 255.255.255.224 ! interface Vlan200 ip address 192.168.250.36 255.255.255.224 ! interface Loopback0 ip address 192.168.5.3 255.255.255.255 ! interface GigabitEthernet1/1 no ip address switchport switchport access vlan 100 switchport mode access ! interface GigabitEthernet1/2 no ip address switchport switchport access vlan 200 switchport mode access ! Distribution – Switch 2 ! vlan 100 name AJ_Lab_Core_Vlan_A ! vlan 200 name AJ_Lab_Core_Vlan_B ! interface Vlan100 ip address 192.168.250.5 255.255.255.224 ! interface Vlan200 ip address 192.168.250.37 255.255.255.224 ! interface Loopback0 ip address 192.168.5.4 255.255.255.255 ! interface GigabitEthernet6/1 switchport switchport access vlan 200 switchport mode access storm-control broadcast level 0.50 ! interface GigabitEthernet6/2 switchport switchport access vlan 100 switchport mode access storm-control broadcast level 0.50 ! Once the distribution layer switches are advertised in core VLANs, the core and distribution layer should establish reachability. Ping check is performed between distribution pair and the core to ensure reachability. Ping checks for core interfaces in VLAN 100 from DS1 and DS2: 30 DS1#ping 192.168.250.2 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.2, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms DS1# DS1#ping 192.168.250.3 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.3, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms DS1# DS2#ping 192.168.250.2 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.2, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms DS2#ping 192.168.250.3 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.3, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms DS2# Ping checks for core interfaces in VLAN 200 from DS1 and DS2: DS1#ping 192.168.250.34 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.34, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms DS1# DS1#ping 192.168.250.35 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.35, timeout is 2 seconds: !!!!! 31 Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms DS1# DS2#ping 192.168.250.34 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.34, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/1/4 ms DS2#ping 192.168.250.35 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.35, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms DS2# The above test ensures that the connectivity is established between the core and the distribution layer and that the distribution layer has redundant connections to the core. However, the end-to-end reachability from access layer to core is not yet established. To establish end-to-end connectivity, the next step is to advertise the access VLAN into core. This can be done via static or dynamic routing. In an enterprise LAN environment, as there are hundreds of access VLAN’s, static routes is not a scalable solution to implement. It will be laborious and hard to manage hundreds of static routes. The feasible solution would be to configure a dynamic routing protocol. 2.7 Configuring EIGRP Enhanced Interior Gateway Routing Protocol (EIGRP) is a dynamic routing protocol developed by Cisco. EIGRP scales effectively in a well-designed network and provides extremely quick convergence times with minimal network traffic [4]. An EIGRP process is identified by an autonomous system (AS) number so that routers with the same AS numbers will exchange routing information with each other, resulting 32 in a routing domain. As the network prototype is implemented with two sites A and B, these sites are assigned with unique AS numbers 10 and 20. Within the EIGRP process, only the dedicated networks are advertised. Like any other routing protocol, EIGRP uses route metric (Bandwidth, Delay, Reliability, Load and MTU) to select the best route form a group of possible routes. The route metrics configured are Bandwidth= 100000, Delay = 100, Reliability=255, Load=1 and MTU=1500. The core and the distribution routers of Site1 are configured as follows: Core-R1#sh run | b eigrp router eigrp 10 network 192.168.5.1 0.0.0.0 network 192.168.250.0 0.0.0.31 network 192.168.250.32 0.0.0.31 network 192.168.250.64 0.0.0.31 default-metric 100000 100 255 1 1500 no auto-summary DS1#sh run | b eigrp router eigrp 10 passive-interface default no passive-interface Vlan100 no passive-interface Vlan200 network 192.168.1.0 network 192.168.3.0 network 192.168.5.3 0.0.0.0 network 192.168.250.0 0.0.0.31 network 192.168.250.32 0.0.0.31 network 192.168.250.64 0.0.0.31 no auto-summary eigrp log-neighbor-changes ! Core-R2#sh run | b eigrp router eigrp 10 network 192.168.5.2 0.0.0.0 network 192.168.250.0 0.0.0.31 network 192.168.250.32 0.0.0.31 network 192.168.250.64 0.0.0.31 default-metric 100000 100 255 1 1500 no auto-summary ! DS2#sh run | b eigrp router eigrp 10 passive-interface default no passive-interface Vlan100 no passive-interface Vlan200 network 192.168.1.0 network 192.168.3.0 network 192.168.5.4 0.0.0.0 network 192.168.250.0 0.0.0.31 network 192.168.250.32 0.0.0.31 network 192.168.250.64 0.0.0.31 no auto-summary eigrp log-neighbor-changes When the devices are configured and the networks are advertised in the EIGRP process, as a part of neighbor discovery process, the EIGRP routers sent hello packets to their 33 connected networks using multicast address 224.0.0.10. The figure 2-8 shows the convergence process between two EIGRP routers [5]. Figure 2-8 Discovering Routes The following messages were captured on DS1 during the neighbor discovery process. DS1 becomes adjacent with Core1, Core2 and DS2 via VLAN100 through subnet 192.168.250.0/27 and also via VLAN 200 through subnet 192.168.250.32/27. 10w4d: %DUAL-5-NBRCHANGE: IP-EIGRP 10: Neighbor 192.168.250.2 (Vlan100) is up: new adjacency 10w4d: %DUAL-5-NBRCHANGE: IP-EIGRP 10: Neighbor 192.168.250.35 (Vlan200) is up: new adjacency 10w4d: %DUAL-5-NBRCHANGE: IP-EIGRP 10: Neighbor 192.168.250.3 (Vlan100) is up: new adjacency 10w4d: %DUAL-5-NBRCHANGE: IP-EIGRP 10: Neighbor 192.168.250.5 (Vlan100) is up: new adjacency 10w4d: %DUAL-5-NBRCHANGE: IP-EIGRP 10: Neighbor 192.168.250.37 (Vlan200) is up: new adjacency 10w4d: %DUAL-5-NBRCHANGE: IP-EIGRP 10: Neighbor 192.168.250.34 34 (Vlan200) is up: new adjacency Once a neighbor is discovered via the hello packet, a neighbor table is created which keeps the state information of the adjacent neighbors. When newly discovered neighbors are learned, the address and the interface of the neighbors are recorded. The following table gives the neighbor table of Core1, Core2, DS1 and DS2. Core-R1#sh ip eigrp neighbors IP-EIGRP neighbors for process 10 H Address Interface Hold Uptime SRTT RTO Q Seq Type (sec) (ms) Cnt Num 6 192.168.250.35 Fa2/0 14 17:23:52 4 200 0 80 4 192.168.250.37 Fa2/0 13 17:23:52 3 200 0 28 1 192.168.250.36 Fa2/0 10 17:23:54 1 200 0 24 2 192.168.250.5 Fa1/0.1 11 17:59:56 1 200 0 29 0 192.168.250.4 Fa1/0.1 12 18:00:50 1 200 0 23 5 192.168.250.3 Fa1/0.1 11 3d00h 1 200 0 79 Core-R1# Core-R2#sh ip eigrp neighbors IP-EIGRP neighbors for process 10 H Address Interface Hold Uptime SRTT RTO Q Seq Type (sec) (ms) Cnt Num 2 192.168.250.34 Fa2/0 13 17:24:22 1568 5000 0 695 6 192.168.250.37 Fa2/0 10 18:00:26 4 200 0 28 4 192.168.250.5 Fa1/0.1 14 18:00:26 2 200 0 29 1 192.168.250.36 Fa2/0 13 18:01:20 4 200 0 24 0 192.168.250.4 Fa1/0.1 10 18:01:20 1 200 0 23 5 192.168.250.2 Fa1/0.1 12 3d00h 37 222 0 696 Core-R2# DS1#sh ip eigrp nei IP-EIGRP neighbors for process 10 H Address Interface Hold Uptime SRTT RTO Q Seq Type (sec) (ms) Cnt Num 0 192.168.250.34 Vl200 10 17:14:21 57 342 0 695 5 192.168.250.37 Vl200 13 17:50:25 3 200 0 28 4 192.168.250.5 Vl100 14 17:50:25 2 200 0 29 3 192.168.250.3 Vl100 12 17:51:19 490 2940 0 79 2 192.168.250.35 Vl200 14 17:51:20 868 5000 0 80 1 192.168.250.2 Vl100 12 17:51:20 316 1896 0 696 DS1# DS2#sh ip eigrp neighbors 35 IP-EIGRP neighbors for process 10 H Address Interface Hold Uptime SRTT RTO Q Seq (sec) (ms) Cnt Num 1 192.168.250.34 Vl200 13 17:24:56 1053 5000 0 695 5 192.168.250.2 Vl100 12 18:00:58 1 200 0 696 4 192.168.250.35 Vl200 14 18:00:58 1 200 0 80 3 192.168.250.3 Vl100 11 18:00:58 1 200 0 79 2 192.168.250.4 Vl100 11 18:00:59 224 1344 0 23 0 192.168.250.36 Vl200 13 18:01:00 454 2724 0 24 DS2# Once the EIGPR neighbor table is created, the routers send an update containing all of the routes that they know about to the neighbor. This information is stored in the topology table. The following table shows the topology table of the EIGRP routers. Core-R1#sh ip eigrp topology IP-EIGRP Topology Table for AS(10)/ID(192.168.5.1) Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply, r - reply Status, s - sia Status P 192.168.250.128/27, 1 successors, FD is 51200, tag is 200 via Redistributed (51200/0) P 192.168.5.4/32, 2 successors, FD is 156160 via 192.168.250.37 (156160/128256), FastEthernet2/0 via 192.168.250.5 (156160/128256), FastEthernet1/0.1 P 192.168.1.0/24, 4 successors, FD is 28416 via 192.168.250.37 (28416/2816), FastEthernet2/0 via 192.168.250.5 (28416/2816), FastEthernet1/0.1 via 192.168.250.4 (28416/2816), FastEthernet1/0.1 via 192.168.250.36 (28416/2816), FastEthernet2/0 P 192.168.3.0/24, 4 successors, FD is 28416 via 192.168.250.37 (28416/2816), FastEthernet2/0 via 192.168.250.5 (28416/2816), FastEthernet1/0.1 via 192.168.250.4 (28416/2816), FastEthernet1/0.1 via 192.168.250.36 (28416/2816), FastEthernet2/0 P 192.168.5.1/32, 1 successors, FD is 128256 via Connected, Loopback0 P 192.168.5.3/32, 2 successors, FD is 156160 via 192.168.250.36 (156160/128256), FastEthernet2/0 via 192.168.250.4 (156160/128256), FastEthernet1/0.1 36 P 192.168.5.2/32, 3 successors, FD is 156160 via 192.168.250.35 (156160/128256), FastEthernet2/0 via 192.168.250.69 (156160/128256), FastEthernet1/0.2 via 192.168.250.3 (156160/128256), FastEthernet1/0.1 P 192.168.250.0/27, 1 successors, FD is 28160 via Connected, FastEthernet1/0.1 via 192.168.250.37 (28416/2816), FastEthernet2/0 via 192.168.250.36 (28416/2816), FastEthernet2/0 P 192.168.250.32/27, 1 successors, FD is 28160 via Connected, FastEthernet2/0 via 192.168.250.4 (28416/2816), FastEthernet1/0.1 via 192.168.250.5 (28416/2816), FastEthernet1/0.1 P 172.16.1.8/30, 3 successors, FD is 53760 via 192.168.250.35 (53760/51200), FastEthernet2/0 via 192.168.250.3 (53760/51200), FastEthernet1/0.1 via 192.168.250.69 (53760/51200), FastEthernet1/0.2 P 172.16.1.4/30, 1 successors, FD is 51200 via Redistributed (51200/0) P 192.168.250.64/27, 1 successors, FD is 28160 via Connected, FastEthernet1/0.2 P 192.168.250.96/27, 1 successors, FD is 51200, tag is 200 Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply, r - reply Status, s - sia Status via Redistributed (51200/0) Core-R1# Core-R2#sh ip eigrp topology IP-EIGRP Topology Table for AS(10)/ID(192.168.5.2) Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply, r - reply Status, s - sia Status P 192.168.250.128/27, 3 successors, FD is 53760, tag is 200 via 192.168.250.34 (53760/51200), FastEthernet2/0 via 192.168.250.2 (53760/51200), FastEthernet1/0.1 via 192.168.250.68 (53760/51200), FastEthernet1/0.2 P 192.168.5.4/32, 2 successors, FD is 156160 via 192.168.250.37 (156160/128256), FastEthernet2/0 via 192.168.250.5 (156160/128256), FastEthernet1/0.1 P 192.168.1.0/24, 4 successors, FD is 28416 via 192.168.250.37 (28416/2816), FastEthernet2/0 via 192.168.250.36 (28416/2816), FastEthernet2/0 via 192.168.250.4 (28416/2816), FastEthernet1/0.1 37 via 192.168.250.5 (28416/2816), FastEthernet1/0.1 P 192.168.3.0/24, 4 successors, FD is 28416 via 192.168.250.37 (28416/2816), FastEthernet2/0 via 192.168.250.36 (28416/2816), FastEthernet2/0 via 192.168.250.4 (28416/2816), FastEthernet1/0.1 via 192.168.250.5 (28416/2816), FastEthernet1/0.1 P 192.168.5.1/32, 3 successors, FD is 156160 Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply, r - reply Status, s - sia Status via 192.168.250.34 (156160/128256), FastEthernet2/0 via 192.168.250.2 (156160/128256), FastEthernet1/0.1 via 192.168.250.68 (156160/128256), FastEthernet1/0.2 P 192.168.5.3/32, 2 successors, FD is 156160 via 192.168.250.36 (156160/128256), FastEthernet2/0 via 192.168.250.4 (156160/128256), FastEthernet1/0.1 P 192.168.5.2/32, 1 successors, FD is 128256 via Connected, Loopback0 P 192.168.250.0/27, 1 successors, FD is 28160 via Connected, FastEthernet1/0.1 via 192.168.250.37 (28416/2816), FastEthernet2/0 via 192.168.250.36 (28416/2816), FastEthernet2/0 P 192.168.250.32/27, 1 successors, FD is 28160 via Connected, FastEthernet2/0 via 192.168.250.5 (28416/2816), FastEthernet1/0.1 via 192.168.250.4 (28416/2816), FastEthernet1/0.1 P 172.16.1.8/30, 1 successors, FD is 51200 via Redistributed (51200/0) P 172.16.1.4/30, 3 successors, FD is 53760 via 192.168.250.34 (53760/51200), FastEthernet2/0 via 192.168.250.2 (53760/51200), FastEthernet1/0.1 via 192.168.250.68 (53760/51200), FastEthernet1/0.2 P 192.168.250.64/27, 1 successors, FD is 28160 Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply, r - reply Status, s - sia Status via Connected, FastEthernet1/0.2 P 192.168.250.96/27, 3 successors, FD is 53760, tag is 200 via 192.168.250.34 (53760/51200), FastEthernet2/0 via 192.168.250.2 (53760/51200), FastEthernet1/0.1 via 192.168.250.68 (53760/51200), FastEthernet1/0.2 Core-R2# 38 DS1#sh ip eigrp topology IP-EIGRP Topology Table for AS(10)/ID(192.168.5.3) Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply, r - reply Status, s - sia Status P 192.168.250.128/27, 2 successors, FD is 51456, tag is 200 via 192.168.250.2 (51456/51200), Vlan100 via 192.168.250.34 (51456/51200), Vlan200 P 192.168.5.4/32, 2 successors, FD is 130816 via 192.168.250.37 (130816/128256), Vlan200 via 192.168.250.5 (130816/128256), Vlan100 P 192.168.1.0/24, 1 successors, FD is 2816 via Connected, Vlan50 P 192.168.3.0/24, 1 successors, FD is 2816 via Connected, Vlan10 P 192.168.5.1/32, 2 successors, FD is 130816 via 192.168.250.2 (130816/128256), Vlan100 via 192.168.250.34 (130816/128256), Vlan200 P 192.168.5.3/32, 1 successors, FD is 128256 via Connected, Loopback0 P 192.168.5.2/32, 2 successors, FD is 130816 via 192.168.250.35 (130816/128256), Vlan200 via 192.168.250.3 (130816/128256), Vlan100 Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply, r - reply Status, s - sia Status P 192.168.250.0/27, 1 successors, FD is 2816 via Connected, Vlan100 P 192.168.250.32/27, 1 successors, FD is 2816 via Connected, Vlan200 P 172.16.1.8/30, 2 successors, FD is 51456 via 192.168.250.35 (51456/51200), Vlan200 via 192.168.250.3 (51456/51200), Vlan100 P 172.16.1.4/30, 2 successors, FD is 51456 via 192.168.250.2 (51456/51200), Vlan100 via 192.168.250.34 (51456/51200), Vlan200 P 192.168.250.64/27, 4 successors, FD is 28416 via 192.168.250.34 (28416/28160), Vlan200 via 192.168.250.2 (28416/28160), Vlan100 via 192.168.250.3 (28416/28160), Vlan100 via 192.168.250.35 (28416/28160), Vlan200 P 192.168.250.96/27, 2 successors, FD is 51456, tag is 200 39 via 192.168.250.2 (51456/51200), Vlan100 via 192.168.250.34 (51456/51200), Vlan200 DS1# DS2#sh ip eigrp topology IP-EIGRP Topology Table for AS(10)/ID(192.168.5.4) Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply, r - reply Status, s - sia Status P 192.168.250.128/27, 2 successors, FD is 51456, tag is 200 via 192.168.250.2 (51456/51200), Vlan100 via 192.168.250.34 (51456/51200), Vlan200 P 192.168.5.4/32, 1 successors, FD is 128256 via Connected, Loopback0 P 192.168.1.0/24, 1 successors, FD is 2816 via Connected, Vlan50 P 192.168.3.0/24, 1 successors, FD is 2816 via Connected, Vlan10 P 192.168.5.1/32, 2 successors, FD is 130816 via 192.168.250.2 (130816/128256), Vlan100 via 192.168.250.34 (130816/128256), Vlan200 P 192.168.5.3/32, 2 successors, FD is 130816 via 192.168.250.4 (130816/128256), Vlan100 via 192.168.250.36 (130816/128256), Vlan200 P 192.168.5.2/32, 2 successors, FD is 130816 via 192.168.250.3 (130816/128256), Vlan100 via 192.168.250.35 (130816/128256), Vlan200 Codes: P - Passive, A - Active, U - Update, Q - Query, R - Reply, r - reply Status, s - sia Status P 192.168.250.0/27, 1 successors, FD is 2816 via Connected, Vlan100 P 192.168.250.32/27, 1 successors, FD is 2816 via Connected, Vlan200 P 172.16.1.8/30, 2 successors, FD is 51456 via 192.168.250.3 (51456/51200), Vlan100 via 192.168.250.35 (51456/51200), Vlan200 P 172.16.1.4/30, 2 successors, FD is 51456 via 192.168.250.2 (51456/51200), Vlan100 via 192.168.250.34 (51456/51200), Vlan200 P 192.168.250.64/27, 4 successors, FD is 28416 via 192.168.250.2 (28416/28160), Vlan100 via 192.168.250.3 (28416/28160), Vlan100 40 via 192.168.250.34 (28416/28160), Vlan200 via 192.168.250.35 (28416/28160), Vlan200 P 192.168.250.96/27, 2 successors, FD is 51456, tag is 200 via 192.168.250.2 (51456/51200), Vlan100 via 192.168.250.34 (51456/51200), Vlan200 DS2# After the EIGRP routers get fully converged and the routing information is exchanged, the access layer devices should now be able to reach the core layer devices. The following table provides the routing table of one the core routers. The access layer routes 192.168.1.0 and 192.168.3.0 are shown in the routing table of Core-R1 that are learnt from EIGRP neighbors and distribution switches 1 and 2 through VLAN’s 100 and 200. Core-R1#sh ip ro Codes: C - connected, S - static, I - IGRP, R - RIP, M - mobile, B - BGP D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2 E1 - OSPF external type 1, E2 - OSPF external type 2, E - EGP i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2 ia - IS-IS inter area, * - candidate default, U - per-user static route o - ODR, P - periodic downloaded static route Gateway of last resort is not set 172.16.0.0/30 is subnetted, 2 subnets D EX 172.16.1.8 [170/53760] via 192.168.250.69, 21:48:15, FastEthernet1/0.2 [170/53760] via 192.168.250.3, 21:48:15, FastEthernet1/0.1 [170/53760] via 192.168.250.35, 21:48:15, FastEthernet2/0 C 172.16.1.4 is directly connected, Serial0/2 192.168.250.0/27 is subnetted, 5 subnets B 192.168.250.128 [20/0] via 172.16.1.6, 7w0d C 192.168.250.0 is directly connected, FastEthernet1/0.1 C 192.168.250.32 is directly connected, FastEthernet2/0 C 192.168.250.64 is directly connected, FastEthernet1/0.2 B 192.168.250.96 [20/0] via 172.16.1.6, 7w0d 192.168.5.0/32 is subnetted, 4 subnets D 192.168.5.4 [90/156160] via 192.168.250.5, 21:48:14, FastEthernet1/0.1 [90/156160] via 192.168.250.37, 21:48:14, FastEthernet2/0 C 192.168.5.1 is directly connected, Loopback0 D 192.168.5.3 [90/156160] via 192.168.250.4, 21:48:19, FastEthernet1/0.1 41 [90/156160] via 192.168.250.36, 21:48:19, FastEthernet2/0 D 192.168.5.2 [90/156160] via 192.168.250.69, 21:48:19, FastEthernet1/0.2 [90/156160] via 192.168.250.3, 21:48:19, FastEthernet1/0.1 [90/156160] via 192.168.250.35, 21:48:19, FastEthernet2/0 D 192.168.1.0/24 [90/28416] via 192.168.250.4, 21:48:17, FastEthernet1/0.1 [90/28416] via 192.168.250.5, 21:48:17, FastEthernet1/0.1 [90/28416] via 192.168.250.36, 21:48:17, FastEthernet2/0 [90/28416] via 192.168.250.37, 21:48:17, FastEthernet2/0 D 192.168.3.0/24 [90/28416] via 192.168.250.4, 21:48:17, FastEthernet1/0.1 [90/28416] via 192.168.250.5, 21:48:17, FastEthernet1/0.1 [90/28416] via 192.168.250.36, 21:48:19, FastEthernet2/0 [90/28416] via 192.168.250.37, 21:48:19, FastEthernet2/0 Core-R1# A ping test is performed from access layer switch to the core routers to ensure end-to-end connectivity. Ping to core router interfaces in VLAN 100 SW2#ping 192.168.250.2 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.2, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/3/12 ms SW2# SW2#ping 192.168.250.3 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.3, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/3/4 ms SW2# Ping to core router interfaces in VLAN 200 SW2#ping 192.168.250.34 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.34, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/4/8 ms 42 SW2#ping 192.168.250.35 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.35, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/4/8 ms SW2# 2.8 Implementation Of Core Layer at Site2 To maintain consistency Site2 was designed similar to Site1 in hierarchical layers. The core layer comprises of two routers and a core switch to connect the LAN distribution layer. A VLAN 300 with subnet 192.168.250.96/27 is configured as a core VLAN for Site2. Figure 2.9 shows the layer2 connectivity of the core devices. Fa2/0 Fa2/0.1 .98 Lo=192.168.5.6 Lo=192.168.5.5 Fa2/0 Fa2/0.1 .99 Site2-Core B 3640 Site2-Core A 3640 Access VLAN200 192.168.250.32/27 Figure 2-9 Site2 Core VLAN EIGRP is enabled between these two core routers and the core subnet is advertised in a site specific EIGRP process 20. The following output is captured from the core routers displaying the EIGRP neighbor relationship. R5#sh ip eigrp nei IP-EIGRP neighbors for process 20 H Address Interface Hold Uptime SRTT RTO Q Seq Type (sec) (ms) Cnt Num 0 192.168.250.99 Fa2/0.1 14 4w2d 1 200 0 57 43 R5# R2#sh ip eigrp nei IP-EIGRP neighbors for process 20 H Address Interface Hold Uptime SRTT RTO Q Seq Type (sec) (ms) Cnt Num 0 192.168.250.98 Fa2/0.1 11 4w2d 1 200 0 55 R2# The site 2 was implemented with the scope of deploying TACACS+ servers at the core layer switch. Hence, the VLAN 200 with /27 subnet will be open for future integration of distribution layer through EIGRP process 20. With the successful implementation of Site1 and Site2, a hierarchal network design is maintained with a future scope of deploying and extending the entire sites. As the goal of the project is to deploy a centralized TACACS+ environment by deploying a centralized Master and regional Slave servers, the two regional sites, 1 and 2, need to be integrated with WAN infrastructure. 2.9 Integration Of Site1 and Site2 Using WAN Links In order to keep a simple WAN design, point-to-point serial links between core routers were integrated. /30 subnets were used from the major 172.16.0.0/12 block which is reserved for WAN infrastructure design. A BGP AS 100 is allocated to Site 1 and a BGP AS 200 is allocated to Site2. The figure 2-10 shows the WAN design of the prototype. In order to achieve end-to-end connectivity between two different sites mutual redistribution of BGP and EIGRP between the sites is created. 44 SITE 1 SITE 2 Site1 – Core1 Site2 – Core1 S0/2 Lo=192.168.5.1 Fa3/0 172.16.1.4/30 S0/2 .6 .5 Fa3/0 .1 BGP 100 172.16.1.12/30 .2 Fa3/0 .14 S0/2 Lo=192.168.5.2 .13 BGP 200 172.16.1.0/30 Fa3/0 Lo=192.168.5.5 .9 Site1 – Core2 172.16.1.8/30 S0/2 .10 Lo=192.168.5.6 Site2 – Core2 Figure 2-10 WAN Connectivity between Site1 and Site2 A Fast Ethernet high availability core links between the site1 routers and also between the site2 routers are established. WAN links between the core routers are configured as follows. A /30 subnet 172.16.1.4/30 is used for the link between Site1-Core1 and Site2-Core1 and a /30 subnet 172.16.1.0/30 is used for the high availability link between site1 routers. Site1-Core1 ! interface Serial0/2 ip address 172.16.1.5 255.255.255.252 no shut ! ! interface FastEthernet3/0 ip address 172.16.1.1 255.255.255.252 duplex auto speed auto ! Site2-Core1 ! interface Serial0/2 ip address 172.16.1.6 255.255.255.252 no shut ! 45 The interfaces should be UP and the neighbors should be discovered. The following output provides the interface status of Site1-Core1 and Site2-Core1. Site1-Core1#sh ip int br | i up Serial0/2 172.16.1.5 YES NVRAM up FastEthernet1/0 unassigned YES NVRAM up FastEthernet1/0.1 192.168.250.2 YES NVRAM up FastEthernet1/0.2 192.168.250.68 YES NVRAM up FastEthernet2/0 192.168.250.34 YES NVRAM up FastEthernet3/0 172.16.1.1 YES NVRAM up Loopback0 192.168.5.1 YES NVRAM up Site1-Core1# up up up up up up up Site1-Core1#sh cdp nei Capability Codes: R - Router, T - Trans Bridge, B - Source Route Bridge S - Switch, H - Host, I - IGMP, r - Repeater Device ID Local Intrfce Core-SW2 Fas 2/0 Core-SW1 Fas 1/0 Site2-Core1 Ser 0/2 Site1-Core2 Fas 3/0 Site1-Core1# Holdtme Capability Platform Port ID 174 SI WS-C3550-2Fas 0/3 135 SI WS-C3560G-Gig 0/3 141 R 3640 Ser 0/2 141 R 3640 Fas 3/0 Site2-Core1#sh ip int br | i up Serial0/2 172.16.1.6 YES NVRAM up FastEthernet2/0 unassigned YES NVRAM up FastEthernet2/0.1 192.168.250.98 YES NVRAM up FastEthernet2/0.2 192.168.250.130 YES NVRAM up FastEthernet3/0 172.16.1.13 YES NVRAM up Loopback0 192.168.5.5 YES NVRAM up Site2-Core1# up up up up up up Site2-Core1#sh cdp nei Capability Codes: R - Router, T - Trans Bridge, B - Source Route Bridge S - Switch, H - Host, I - IGMP, r - Repeater Device ID Local Intrfce Site2-Core-SW1 Fas 2/0 Site1-Core1 Ser 0/2 Site2-Core1# Holdtme Capability Platform Port ID 123 SI WS-C3560G-Gig 0/1 148 R 3640 Ser 0/2 46 Ping the far end interfaces to ensure the devices are reachable. Site1-Core1#ping 172.16.1.6 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 172.16.1.6, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/3/4 ms Site1-Core1# Site2-Core1#ping 172.16.1.5 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 172.16.1.5, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 4/4/4 ms Site2-Core1# Similarly the link between Site1-Core2 and Site2-Core2 is established using the subnet 172.16.1.8/30. The following configuration is applied on the routers. Site1-Core2 ! interface Serial0/2 ip address 172.16.1.9 255.255.255.252 ! interface FastEthernet3/0 ip address 172.16.1.2 255.255.255.252 duplex auto speed auto ! Site2-Core2 ! interface Serial0/2 ip address 172.16.1.10 255.255.255.252 ! interface FastEthernet3/0 ip address 172.16.1.14 255.255.255.252 duplex auto speed auto ! The interfaces should be UP and the neighbors should be discovered. The following output provides the interface status of Site1-Core2 and Site2-Core2. Site1-Core2#sh ip int br | i up Serial0/2 172.16.1.9 YES NVRAM up FastEthernet1/0 unassigned YES NVRAM up FastEthernet1/0.1 192.168.250.3 YES NVRAM up FastEthernet1/0.2 192.168.250.69 YES NVRAM up FastEthernet2/0 192.168.250.35 YES NVRAM up FastEthernet3/0 172.16.1.2 YES NVRAM up up up up up up up 47 Loopback0 Site1-Core2# 192.168.5.2 YES NVRAM up up Site1-Core2#sh cdp nei Capability Codes: R - Router, T - Trans Bridge, B - Source Route Bridge S - Switch, H - Host, I - IGMP, r - Repeater Device ID Core-SW2 Core-SW1 Site2-Core2 Site1-Core2 Local Intrfce Fas 2/0 Fas 1/0 Ser 0/2 Fas 3/0 Holdtme Capability Platform Port ID 139 SI WS-C3550-2Fas 0/4 161 SI WS-C3560G-Gig 0/4 171 R 3640 Ser 0/2 141 R 3640 Fas 3/0 Site1-Core2# Site2-Core2#sh ip int br | i up Serial0/2 172.16.1.10 YES NVRAM up FastEthernet2/0 unassigned YES NVRAM up FastEthernet2/0.1 192.168.250.99 YES NVRAM up FastEthernet2/0.2 192.168.250.131 YES NVRAM up FastEthernet3/0 172.16.1.14 YES NVRAM up Loopback0 192.168.5.6 YES NVRAM up Site2-Core2# up up up up up up Ping the far end interfaces to ensure the devices are reachable. Site1-Core2#ping 172.16.1.10 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 172.16.1.10, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/3/4 ms Site1-Core2# Site2-Core2#ping 172.16.1.9 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 172.16.1.9, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/3/4 ms Site2-Core2# 48 The above test ensures that the serial links between the two regional sites are established. BGP can be configured on the core routers by advertising the WAN subnets. The following BGP configuration is applied on the core routers. Site1-Core1 ! router bgp 100 no synchronization bgp log-neighbor-changes network 172.16.1.0 mask 255.255.255.252 network 172.16.1.4 mask 255.255.255.252 neighbor 172.16.1.2 remote-as 100 neighbor 172.16.1.6 remote-as 200 no auto-summary ! Site1-Core2 ! router bgp 100 no synchronization bgp log-neighbor-changes network 172.16.1.0 mask 255.255.255.252 network 172.16.1.8 mask 255.255.255.252 neighbor 172.16.1.1 remote-as 100 neighbor 172.16.1.10 remote-as 200 no auto-summary ! Site2-Core1 ! router bgp 200 no synchronization bgp log-neighbor-changes network 172.16.1.4 mask 255.255.255.252 network 172.16.1.12 mask 255.255.255.252 neighbor 172.16.1.5 remote-as 100 neighbor 172.16.1.14 remote-as 200 no auto-summary ! Site2-Core2 ! router bgp 200 no synchronization bgp log-neighbor-changes network 172.16.1.8 mask 255.255.255.252 network 172.16.1.12 mask 255.255.255.252 neighbor 172.16.1.9 remote-as 100 neighbor 172.16.1.13 remote-as 200 no auto-summary ! The following check ensures that the BGP neighbors are established showing their respective autonomous system. Site1-Core1#sh ip bgp sum BGP router identifier 192.168.5.1, local AS number 100 BGP table version is 297, main routing table version 297 13 network entries using 1261 bytes of memory 13 path entries using 468 bytes of memory 6 BGP path attribute entries using 360 bytes of memory 1 BGP AS-PATH entries using 24 bytes of memory 49 0 BGP route-map cache entries using 0 bytes of memory 0 BGP filter-list cache entries using 0 bytes of memory BGP using 2113 total bytes of memory BGP activity 65/52 prefixes, 288/275 paths, scan interval 60 secs Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd 172.16.1.2 4 100 63161 63200 0 0 0 1w0d 2 172.16.1.6 4 200 108736 108809 297 0 0 10w5d 2 Site1-Core1# Site2-Core1#sh ip bgp sum BGP router identifier 192.168.5.5, local AS number 200 BGP table version is 150, main routing table version 150 12 network entries using 1164 bytes of memory 17 path entries using 612 bytes of memory 8 BGP path attribute entries using 480 bytes of memory 1 BGP AS-PATH entries using 24 bytes of memory 0 BGP route-map cache entries using 0 bytes of memory 0 BGP filter-list cache entries using 0 bytes of memory BGP using 2280 total bytes of memory BGP activity 39/27 prefixes, 129/112 paths, scan interval 60 secs Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd 172.16.1.5 4 100 108817 108742 150 0 0 10w5d 5 172.16.1.14 4 200 108841 108824 150 0 0 10w5d 5 Site2-Core1# Site1-Core2#sh ip bgp sum BGP router identifier 192.168.5.2, local AS number 100 BGP table version is 55, main routing table version 55 13 network entries using 1261 bytes of memory 15 path entries using 540 bytes of memory 6 BGP path attribute entries using 360 bytes of memory 1 BGP AS-PATH entries using 24 bytes of memory 0 BGP route-map cache entries using 0 bytes of memory 0 BGP filter-list cache entries using 0 bytes of memory BGP using 2185 total bytes of memory BGP activity 17/4 prefixes, 21/6 paths, scan interval 60 secs Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd 172.16.1.1 4 100 0 0 0 0 0 never 2 172.16.1.10 4 200 5287 5300 55 0 0 3d16h 2 Site1-Core2# Site2-Core2#sh ip bgp sum BGP router identifier 192.168.5.6, local AS number 200 BGP table version is 163, main routing table version 163 50 12 network entries using 1164 bytes of memory 27 path entries using 972 bytes of memory 10 BGP path attribute entries using 600 bytes of memory 1 BGP AS-PATH entries using 24 bytes of memory 0 BGP route-map cache entries using 0 bytes of memory 0 BGP filter-list cache entries using 0 bytes of memory BGP using 2760 total bytes of memory BGP activity 33/21 prefixes, 121/94 paths, scan interval 60 secs Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down State/PfxRcd 172.16.1.9 4 100 68408 68341 163 0 0 3d16h 5 172.16.1.13 4 200 108758 108781 163 0 0 10w5d 10 Site2-Core2# The serial WAN links between Site1 and Site2 are established and the BGP neighbors are discovered between the sites. However, to establish end-to-end connectivity between the two sites mutual redistribution need to be created between EIGRP and BGP on each site. Figure 2-11 shows an example of redistribution at Site1. In order to avoid routing loops an access-list of only the permitted subnets is created and distributed out of the serial interfaces facing the WAN. 51 Redistribution BGP Into EIGRP Site1 – Core1 S0/2 Lo=192.168.5.1 .5 .1 Fa3/0 BGP 100 EIGRP 10 172.16.1.0/30 Fa3/0 .2 S0/2 Lo=192.168.5.2 .9 Site1 – Core2 Redistribution EIGRP Into BGP Figure 2-11 Redistribution of EIGRP into BGP and BGP into EIGRP at Site1 The following configuration is applied on Site1 and Site2 core routers to create mutual redistribution. Mutual Redistribution on Site1 Core Routers: Site1-Core1 ! access-list 1 permit 192.168.250.0 0.0.0.31 access-list 1 permit 192.168.250.32 0.0.0.31 access-list 1 permit 192.168.250.64 0.0.0.31 access-list 1 permit 192.168.1.0 0.0.0.255 access-list 1 permit 192.168.3.0 0.0.0.255 ! router eigrp 10 Site1-Core2 ! access-list 1 permit 192.168.250.0 0.0.0.31 access-list 1 permit 192.168.250.32 0.0.0.31 access-list 1 permit 192.168.250.64 0.0.0.31 access-list 1 permit 192.168.1.0 0.0.0.255 access-list 1 permit 192.168.3.0 0.0.0.255 ! 52 redistribute bgp 100 ! router bgp 100 redistribute eigrp 10 neighbor 172.16.1.6 distribute-list 1 out ! router eigrp 10 redistribute bgp 100 ! router bgp 100 redistribute eigrp 10 neighbor 172.16.1.10 distribute-list 1 out ! Mutual Redistribution on Site2 core routers: Site2-Core1 ! access-list 1 permit 192.168.250.96 0.0.0.31 access-list 1 permit 192.168.250.128 0.0.0.31 ! ! router eigrp 20 redistribute bgp 200 ! router bgp 200 redistribute eigrp 20 neighbor 172.16.1.5 distribute-list 1 out ! Site2-Core2 ! access-list 1 permit 192.168.250.96 0.0.0.31 access-list 1 permit 192.168.250.128 0.0.0.31 ! ! router eigrp 20 redistribute bgp 200 ! router bgp 200 redistribute eigrp 20 neighbor 172.16.1.9 distribute-list 1 out ! With mutual redistribution, the Site1 LAN devices will learn routes of Site2 LAN devices, and vice versa, as external EIGRP routes. The following output from the distribution layer switches shows the route 192.168.250.96/27 of Site2 core VLAN 300 via VLAN100 and VLAN 200 as a result of redistribution. It also shows the BGP routes as External EIGRP due to redistribution. DS1#sh ip ro Codes: C - connected, S - static, I - IGRP, R - RIP, M - mobile, B - BGP D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2 E1 - OSPF external type 1, E2 - OSPF external type 2, E - EGP i - IS-IS, L1 - IS-IS level-1, L2 - IS-IS level-2, ia - IS-IS inter area * - candidate default, U - per-user static route, o - ODR P - periodic downloaded static route 53 Gateway of last resort is not set 172.16.0.0/30 is subnetted, 2 subnets D EX 172.16.1.8 [170/51456] via 192.168.250.3, 1d11h, Vlan100 [170/51456] via 192.168.250.35, 1d11h, Vlan200 D EX 172.16.1.4 [170/51456] via 192.168.250.2, 1d10h, Vlan100 [170/51456] via 192.168.250.34, 1d10h, Vlan200 192.168.250.0/27 is subnetted, 5 subnets D EX 192.168.250.128 [170/51456] via 192.168.250.2, 1d10h, Vlan100 [170/51456] via 192.168.250.34, 1d10h, Vlan200 C 192.168.250.0 is directly connected, Vlan100 C 192.168.250.32 is directly connected, Vlan200 D 192.168.250.64 [90/28416] via 192.168.250.3, 1d10h, Vlan100 [90/28416] via 192.168.250.2, 1d10h, Vlan100 [90/28416] via 192.168.250.35, 1d10h, Vlan200 [90/28416] via 192.168.250.34, 1d10h, Vlan200 D EX 192.168.250.96 [170/51456] via 192.168.250.2, 1d10h, Vlan100 [170/51456] via 192.168.250.34, 1d10h, Vlan200 192.168.5.0/32 is subnetted, 4 subnets DS2#sh ip ro Codes: C - connected, S - static, R - RIP, M - mobile, B - BGP D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2 E1 - OSPF external type 1, E2 - OSPF external type 2 i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2 ia - IS-IS inter area, * - candidate default, U - per-user static route o - ODR, P - periodic downloaded static route Gateway of last resort is not set 172.16.0.0/30 is subnetted, 2 subnets D EX 172.16.1.4 [170/51456] via 192.168.250.34, 1d10h, Vlan200 [170/51456] via 192.168.250.2, 1d10h, Vlan100 D EX 172.16.1.8 [170/51456] via 192.168.250.35, 1d11h, Vlan200 [170/51456] via 192.168.250.3, 1d11h, Vlan100 192.168.250.0/27 is subnetted, 5 subnets C 192.168.250.0 is directly connected, Vlan100 C 192.168.250.32 is directly connected, Vlan200 D 192.168.250.64 [90/28416] via 192.168.250.35, 1d10h, Vlan200 [90/28416] via 192.168.250.34, 1d10h, Vlan200 [90/28416] via 192.168.250.3, 1d10h, Vlan100 [90/28416] via 192.168.250.2, 1d10h, Vlan100 54 D EX 192.168.250.96 [170/51456] via 192.168.250.34, 1d10h, Vlan200 [170/51456] via 192.168.250.2, 1d10h, Vlan100 D EX 192.168.250.128 [170/51456] via 192.168.250.34, 1d10h, Vlan200 [170/51456] via 192.168.250.2, 1d10h, Vlan100 192.168.5.0/32 is subnetted, 4 subnets D 192.168.5.2 [90/130816] via 192.168.250.35, 1d11h, Vlan200 [90/130816] via 192.168.250.3, 1d11h, Vlan100 D 192.168.5.3 [90/130816] via 192.168.250.36, 1d11h, Vlan200 [90/130816] via 192.168.250.4, 1d11h, Vlan100 D 192.168.5.1 [90/130816] via 192.168.250.34, 1d10h, Vlan200 [90/130816] via 192.168.250.2, 1d10h, Vlan100 C 192.168.5.4 is directly connected, Loopback0 C 192.168.1.0/24 is directly connected, Vlan50 C 192.168.3.0/24 is directly connected, Vlan10 DS2# Similarly the core devices of Site2 learnt routers of Site1 via redistribution. Perform a ping check from the access-layer switch in Site1 to the core subnet in Site2 to ensure endto-end connectivity is established. Access-Switch#ping 192.168.250.99 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.99, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 4/5/12 ms SW2# Site2-Core2#ping 192.168.1.4 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.1.4, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 4/8/24 ms Site2-Core2# The ping shows 100% success rate of end-to-end reachability between the two sites. 55 CHAPTER 3. DEPLOYMENT OF CENTRALIZED TACACS+ ENVIRONMENT The second phase of the project is to deploy a centralized TACACS+ environment to provide AAA services to establish network connectivity to the network elements. In order to deploy TACACS+, a Cisco Secure ACS appliance is built over Win2003 server. A centralized dedicated master server is built along with dedicated primary and secondly slaves in each site. At each site a specific VLAN called as Server-VLAN with a /27 subnet is configured. In order to provide quick authentication and also to reduce the communication time between the TACACS+ servers itself, the VLAN’s are connected to core layer switches. With a future scope of deploying more critical applications the server VLANs are configured with /27 subnets. A VLAN 30 with subnet 192.168.250.64/27 is configured at site1 and a VLAN 40 with subnet 192.168.250.128/27 is configured at site2. Both these VLAN are advertised in EIGRP process and are automatically advertised to the remote sites as a result of redistribution. 56 Fa1/0 Fa1/0.1 .2 Fa1/0.2 .68 Lo=192.168.5.1 Site1-Core 1 3640 Lo=192.168.5.2 Site1-Core 2 3640 Fa1/0 Fa1/0.1 .3 Fa1/0.2 .69 VLAN 30 192.168.250.64/27 Eth 3 .70 MASTER Eth0 .71 SLAVE - 1 Figure 3-1 Server VLAN 30 at Site1 Figure 3-1 shows the layer 2 connectivity of server VLAN 30. The layer 3 interfaces of the server VLAN is configured at site1 core routers. As there is only one physical interface on core routers in site 1, router-on-a-stick is implemented where two logical sub-interfaces are created on one physical interface Fa1/0. Sub-interface fa1/0.1 is configured with core VLAN 100 and the second sub-interface fa1/0.2 is configured with server VLAN 30. Similarly, router-on-a-stick is implemented on Core B of Site1. The master and slave servers are configured with one of the IP addresses within the server VLAN. IP address 192.168.250.70 is assigned to master server and IP address 192.168.250.71 is assigned to the first slave server. The core routers are configured as below: Site1-Core1 ! interface FastEthernet1/0.1 Site1-Core2 ! interface FastEthernet1/0.1 57 encapsulation dot1Q 100 ip address 192.168.250.2 255.255.255.224 ! interface FastEthernet1/0.2 encapsulation dot1Q 30 ip address 192.168.250.68 255.255.255.224 ! encapsulation dot1Q 100 ip address 192.168.250.3 255.255.255.224 ! interface FastEthernet1/0.2 encapsulation dot1Q 30 ip address 192.168.250.69 255.255.255.224 ! A trunk is configured between the two core switches to exchange the server VLAN 30. The core switches are configured as allow VLAN 30. Core-SW1 ! interface GigabitEthernet0/3 description connected to site1-core1 switchport trunk encapsulation dot1q switchport trunk allowed vlan 30,100 switchport mode trunk ! interface GigabitEthernet0/4 description connected to site1-core2 switchport trunk encapsulation dot1q switchport trunk allowed vlan 30,100 switchport mode trunk ! interface GigabitEthernet0/5 description connected to Master Server switchport access vlan 30 switchport mode access ! interface GigabitEthernet0/24 description connected to core-sw2 switchport trunk encapsulation dot1q switchport trunk allowed vlan 30 switchport mode trunk ! Core-SW2 ! ! interface FastEthernet0/5 description connected to Slave Server 1 switchport access vlan 30 switchport mode access ! interface FastEthernet0/24 description connected to core-sw1 switchport trunk encapsulation dot1q switchport trunk allowed vlan 30 switchport mode trunk ! The layer3 interface should be reachable from the core-switches. The following ping ensures the connectivity to the layer 3 interfaces. 58 Core-SW1#ping 192.168.250.68 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.68, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/203/1007 ms Core-SW1# Core-SW1#ping 192.168.250.69 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.69, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/203/1007 ms Core-SW1# Core-SW2#ping 192.168.250.68 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.68, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/202/1000 ms Core-SW2#ping 192.168.250.69 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.69, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/201/1000 ms Core-SW2# 3.1 Installing Master And Slave TACACS+ Servers A Win 2003 server is built as a prerequisite to install TACACS+ using the Cisco Secure ACS 4.2 appliance. Certain application features such as disabling security services, installing Java are performed as prerequisite. The installation guide can be found at the Cisco online documentation [6]. Once the installation is complete, the ACS web interface can be accessed with the URL http://127.0.0.1:2002 as shown in figure 3-2. 59 Figure 3-2 ACS Web Interface Several System configuration settings such as Logging, DB Backup and Restore, Database Replication are configured as per the requirement such as the DB is set to backup once a day; keeping all passed authorization logs for last 180 days, etc. 3.2 Building Master Server And Replicating To Slave Server An ACS server, Server2, is configured as a master server to provide centralized administration. All user level privileges such as enable or read-only right to network devices or groups are performed at the master server. Every time the database is updated it is replicated to the slave servers. The master server can be administered to replicate DB at any number of times to over write the slaves and the slaves are used to provide AAA services to the network users. 60 TACACS (AAA) Master (Centralized) ONE WAY REPLICATION TACACS Slave1 Access Request NE1 TACACS Slave2 Access Request Access Response Site1 NE2 NE1 Access Response Site2 NE2 Figure 3-3 Centralized Administration and One Way Replication As shown in figure 3-3, with limited number of servers available in prototype deployment, the TACACS+ environment is designed in such a way that a master and a slave are deployed in site1 and a second slave server is deployed in site2. Slave1 will act as the primary server for site1 network devices and as secondary or backup for site2 network devices. Similarly, Slave2 will act as the primary server for site2 network devices and as a secondary for site1 network devices. 3.2.1 Creating Network Device Groups The figure 3-4 shows a Network Device Group (NDG) of AAA servers. All the slaves including master server are added in a common group and each slave server is provided with a common secret key of the AAA server. For correct operation the key must be 61 identical on the remote AAA server and ACS. If the shared secret does not match, ACS discards all packets from the remote AAA server. Figure 3-4 NDG of AAA Servers A Network Device Group (NDG) is created for each LAN site and an NDG for WAN access. Figure 3-5 Lab NDGs with LAN &WAN Subnets Each group has subnets associated to each site. For example in figure 3-6, site 1 has two IP ranges which are specific to site1 only. Figure 3-6 Site1 IP Range 62 3.2.2 Creating User Groups A user group is nothing but a group with settings common to the users associated to that group. Two user groups Site1_Users and Site2_Users are created as show in figure 3-7. Figure 3-7 User Groups Each user group is configured with setting permitted to the respective NDG’s. For example in figure 3-8, site1_users group is set to allow all the site1 subnets associated to that group so that all the users associated to that group will only have access to Site1 subnets. Similarly, settings such as enable privileges, aging policies are set to each group. Figure 3-8 NAS Restrictions 3.2.3 Creating Users Two usernames AJ and Bob are created and assigned to the user group Site1_Users and Site2_Users respectively. The user AJ will inherit the group’s settings and Bob will 63 inherit the group settings of site2. A password ‘testpass1’ is assigned to AJ with enable pass as ‘ena123’ and a password ‘testpass2’ is assigned to Bob. Using these credentials AJ and Bob can be able to login to any device in Site1 and Site2 respectively. Figure 3-9 shows an example of user name ‘aj’ assigned to the user group Site1_Users. Figure 3-9 Group Level Settings 3.2.4 Replicating Database Now that the server 2 is designed with NDGs, Users and User Groups, server2 can be prepared as master server for replicating the slave servers 1 and 2. 3.2.5 Master Setup Under the database replication setup in system configuration, the replication components are set to “send” to the receiving slaves. Figure 3-10 Replicating Components from Master Server The replicating partners are the slave servers. 64 Figure 3-11 Replication Servers from Master 3.2.6 Slave Setup Under the slave servers the replicating components are set to “receive” from the master. Figure 3-12 Replicating Components Receiving from Master Server And the inbound replication is accepted from the master server. Figure 3-13 Inbound Replication Master Server Once both master and slave AAA servers are configured the ‘replicate now’ is pushed. The complete master database along with all setting will be pushed to the slave servers. The following logs are captured while the database is being replicated to the slaves. 65 Figure 3-14 Outbound Database Replication from Master Server From the log capture in figure 3-14 it can be observed that once the outbound replication cycle is started and each component is replicated to the slave servers fmops-svr1 and fmops-svr3. And simultaneously, the inbound replication cycle started on the slave server. Figure 3-15 shows an inbound database replication from the master server. 66 Figure 3-15 Inbound Database Replication on Slave Server The database replication can be schedule to any number of times and every time a change is made on the master server it gets replicated to the slaves by overwriting the old database. The next step is to configure the network devices and ensure that authentication and authorization are working and that all the administrative logs are being captured on the slave servers. 3.2.7 Configuring Site1 and Site2 Network Devices With TACACS+ The following tacacs+ configuration is applied on the network devices. ! aaa new-model ! aaa authentication login default group tacacs+ line aaa authentication enable default group tacacs+ enable aaa accounting exec default start-stop group tacacs+ aaa accounting commands 1 default stop-only group tacacs+ aaa accounting commands 15 default stop-only group tacacs+ aaa accounting system default start-stop group tacacs+ 67 aaa session-id common tacacs-server host 192.168.250.70 tacacs-server host 192.168.250.132 tacacs-server timeout 30 tacacs-server directed-request tacacs-server key secretkey ! When the configuration is applied on any device, the “aaa new-model” turns on TACACS+ and the “aaa authentication” line says the default way for users to log in is by checking with the group of tacacs+ servers (any TACACS+ servers defined with the "tacacs-server" command). If that fails to respond, the line or enable password can then be used as fallbacks. The accounting commands creates logs into ACS such as who is currently logged into ACS, what commands are executed, etc. In the above configuration the first server in the config will act as the primary server, in this case 192.168.250.70 and the second server in the config will act as the secondary server, in this case 192.168.250.132. The exact same TACACS+ configuration is applied on Site2 network device except that the primary tacacs server is configured as 192.168.250.132 and the secondary server is chosen as 192.168.250.70. tacacs-server host 192.168.250.132 tacacs-server host 192.168.250.70 With the above configuration is applied on a device and a telnet or ssh request is sent, username and password is prompted. Based on the user input user access is either authorized or denied. Figure 3-15 [7] shows a communication between a Client and a TACACS+ Server and the communication involves seven different types of messages. 68 Figure 3-16 Communication between Client and TACACS+ Server TACACS+ Devices the packet or message definations as follows [8]: Authentication START (It describes the type of authentication to be performed, and may contain the username and some authentication data. The START packet is only ever sent as the first message in a TACACS+ authentication session.). Authentication REPLY (It indicates whether the authentication is finished, or whether it should continue. If the REPLY indicates that authentication should continue, then it will also indicate what new information is requested.). Authentication CONTINUE (It is sent from the NAS to the server following the receipt of a REPLY packet and possibly contains requested information.). 69 Authorization REQUEST (It contains a fixed set of fields that describe the authenticity of the user or process, and a variable set of arguments that describes the services and options for which authorization is requested.). Authorization RESPONSE (It contains a variable set of response arguments (attribute-value pairs) which can restrict or modify the client actions.). Accounting REQUEST (It conveys information used to provide accounting for a service provided to a user.). Accounting REPLY (It is used to indicate that the accounting function on the server has completed and securely committed the record.). 3.2.8 Testing For PASSED Authorization To Network Devices The config is applied on one of the distribution layer switches DS2 and the following debugging messages are captured when an access request is sent. Debug tacacs is enabled on DS2. For each command tacacs debugging messages are captured on DS2. A telnet request is initiated from DS1 with the management IP address of DS2. DS1#telnet 192.168.1.2 Trying 192.168.1.2 ... Open AJ-LAB-Test-Bed Please do not delete any configuration Please do not pull any cables Contact AJ - 377-5198 Username: Observed that an authentication request is started using the primary server 192.168.250.70 and an authentication response status of GET_USER is initiated. 70 12w1d: TPLUS: Queuing AAA Authentication request 52 for processing 12w1d: TPLUS: processing authentication start request id 52 12w1d: TPLUS: Authentication start packet created for 52() 12w1d: TPLUS: Using server 192.168.250.70 12w1d: TPLUS(00000034)/0/NB_WAIT/479328E8: Started 30 sec timeout 12w1d: TPLUS(00000034)/0/NB_WAIT: socket event 2 12w1d: TPLUS(00000034)/0/NB_WAIT: wrote entire 35 bytes request 12w1d: TPLUS(00000034)/0/READ: socket event 1 12w1d: TPLUS(00000034)/0/READ: Would block while reading 12w1d: TPLUS(00000034)/0/READ: socket event 1 12w1d: TPLUS(00000034)/0/READ: read entire 12 header bytes (expect 16 bytes data) 12w1d: TPLUS(00000034)/0/READ: socket event 1 12w1d: TPLUS(00000034)/0/READ: read entire 28 bytes response 12w1d: TPLUS(00000034)/0/479328E8: Processing the reply packet 12w1d: TPLUS: Received authen response status GET_USER (7) Username ‘AJ’ is entered on the username prompt. Username: aj A GET_PASSWORD response is initiated. Password: The user AJ’s password ‘testpass1’ is then entered. The authentication request for the given password is being processed. 12w1d: TPLUS: Queuing AAA Authentication request 53 for processing 12w1d: TPLUS: processing authentication continue request id 53 12w1d: TPLUS: Authentication continue packet generated for 53 12w1d: TPLUS(00000035)/0/WRITE/468E2D30: Started 30 sec timeout 12w1d: TPLUS(00000035)/0/WRITE: wrote entire 19 bytes request 12w1d: TPLUS(00000035)/0/READ: socket event 1 12w1d: TPLUS(00000035)/0/READ: read entire 12 header bytes (expect 16 bytes data) 12w1d: TPLUS(00000035)/0/READ: socket event 1 12w1d: TPLUS(00000035)/0/READ: read entire 28 bytes response 12w1d: TPLUS(00000035)/0/468E2D30: Processing the reply packet 12w1d: TPLUS: Received authen response status GET_PASSWORD (8) 71 Once the authentication response is PASS the user is authorized and AAA accounting request is processed. 12w1d: TPLUS: Queuing AAA Authentication request 53 for processing 12w1d: TPLUS: processing authentication continue request id 53 12w1d: TPLUS: Authentication continue packet generated for 53 12w1d: TPLUS(00000035)/0/WRITE/477DECD8: Started 30 sec timeout 12w1d: TPLUS(00000035)/0/WRITE: wrote entire 26 bytes request 12w1d: TPLUS(00000035)/0/READ: socket event 1 12w1d: TPLUS(00000035)/0/READ: read entire 12 header bytes (expect 6 bytes data) 12w1d: TPLUS(00000035)/0/READ: socket event 1 12w1d: TPLUS(00000035)/0/READ: read entire 18 bytes response 12w1d: TPLUS(00000035)/0/477DECD8: Processing the reply packet 12w1d: TPLUS: Received authen response status PASS (2) 12w1d: TPLUS: Queuing AAA Accounting request 53 for processing 12w1d: TPLUS: processing accounting request id 53 12w1d: TPLUS: Sending AV task_id=11 12w1d: TPLUS: Sending AV timezone=UTC 12w1d: TPLUS: Sending AV service=shell 12w1d: TPLUS: Accounting request created for 53(aj) 12w1d: TPLUS: using previously set server 192.168.250.71 from group tacacs+ 12w1d: TPLUS(00000035)/0/NB_WAIT/468E2D30: Started 30 sec timeout 12w1d: TPLUS(00000035)/0/NB_WAIT: socket event 2 12w1d: TPLUS(00000035)/0/NB_WAIT: wrote entire 76 bytes request 12w1d: TPLUS(00000035)/0/READ: socket event 1 12w1d: TPLUS(00000035)/0/READ: Would block while reading 12w1d: TPLUS(00000035)/0/READ: socket event 1 12w1d: TPLUS(00000035)/0/READ: read entire 12 header bytes (expect 5 bytes data) 12w1d: TPLUS(00000035)/0/READ: socket event 1 12w1d: TPLUS(00000035)/0/READ: read entire 17 bytes response 12w1d: TPLUS(00000035)/0/468E2D30: Processing the reply packet 12w1d: TPLUS: Received accounting response with status PASS When the user gets into the enable mode a GETPASS authentication status request is initiated. DS1>enable 12w1d: TAC+: send AUTHEN/START packet ver=192 id=2847203026 12w1d: TAC+: Using default tacacs server-group "tacacs+" list. 12w1d: TAC+: Opening TCP/IP to 192.168.250.71/49 timeout=30 72 12w1d: TAC+: Opened TCP/IP handle 0x522D4EC8 to 192.168.250.70/49 12w1d: TAC+: 192.168.250.71 (2847203026) AUTHEN/START/LOGIN/ASCII queued 12w1d: TAC+: (2847203026) AUTHEN/START/LOGIN/ASCII processed 12w1d: TAC+: ver=192 id=2847203026 received AUTHEN status = GETPASS The user enters the password an AUTHEN/CONT packet is sent to the AAA server for the password to be processed. Finally when the user exits from the device the TCP/IP Connection is closed with the TACACS+ Server. 12w1d: TAC+: send AUTHEN/CONT packet id=2847203026 12w1d: TAC+: 192.168.250.71 (2847203026) AUTHEN/CONT queued 12w1d: TAC+: (2847203026) AUTHEN/CONT processed 12w1d: TAC+: ver=192 id=2847203026 received AUTHEN status = PASS 12w1d: TAC+: Closing TCP/IP 0x522D4EC8 connection to 192.168.250.70/49 Following are the ‘aaa authentication’ debug messages from successful telnet to exec mode on the device. Observe that the user ‘aj’ is permitted with full enable privileges 15. 12w1d: AAA/BIND(00000031): Bind i/f 12w1d: AAA/AUTHEN/LOGIN (00000031): Pick method list 'default' 12w1d: AAA/AUTHEN/LOGIN (00000031): Pick method list 'default' 12w1d: AAA: parse name=tty1 idb type=-1 tty=-1 12w1d: AAA: name=tty1 flags=0x11 type=5 shelf=0 slot=0 adapter=0 port=1 channel=0 12w1d: AAA/MEMORY: create_user (0x522C4140) user='aj' ruser='NULL' ds0=0 port='tty1' rem_addr='192.168.1.3' authen_type=ASCII service=ENABLE priv=15 initial_task_id='0', vrf= (id=0) 12w1d: AAA/AUTHEN/START (828036695): port='tty1' list='' action=LOGIN service=ENABLE 12w1d: AAA/AUTHEN/START (828036695): using "default" list 12w1d: AAA/AUTHEN/START (828036695): Method=tacacs+ (tacacs+) 12w1d: TAC+: send AUTHEN/START packet ver=192 id=828036695 12w1d: TAC+: ver=192 id=828036695 received AUTHEN status = GETPASS 12w1d: AAA/AUTHEN (828036695): status = GETPASS 12w1d: AAA/AUTHEN/CONT (828036695): continue_login (user='aj') 12w1d: AAA/AUTHEN (828036695): status = GETPASS 12w1d: AAA/AUTHEN (828036695): Method=tacacs+ (tacacs+) 12w1d: TAC+: send AUTHEN/CONT packet id=828036695 12w1d: TAC+: ver=192 id=828036695 received AUTHEN status = PASS 12w1d: AAA/AUTHEN (828036695): status = PASS 73 12w1d: AAA/MEMORY: free_user (0x522C4140) user='aj' ruser='NULL' port='tty1' rem_addr='192.168.1.3' authen_type=ASCII service=ENABLE priv=15 Now that the user ‘aj’ is permitted into the device, any command that is executed is logged into the ACS server. Figure 3-17 Authentication Logs of User AJ From the above logs in figure 3-16, it can be observed that all the commands executed by the user AJ are captured. 3.2.9 Testing For FAILED Authorization To Network Devices Using the same credentials of user AJ, a device in Site2 is sent a telnet request to establish connectivity. Since user AJ is configured in ACS to be authorized only to Site1 devices, the authorization should fail. A site2 core switch is sent a telnet request from Site1 distribution layer switch. The following TACACS debug authentication messages are captured on the core switch in site2 and the telnet authorization fails for user AJ. When a telnet request is sent from DS2, the core device Site2-Core-SW1 prompts for username. DS2#telnet 192.168.250.97 Trying 192.168.250.97 ... Open 74 User Access Verification Username: Debug message on Site2-Core-SW1 Site2-Core-SW1# 13w6d: AAA: parse name=tty1 idb type=-1 tty=-1 13w6d: AAA: name=tty1 flags=0x11 type=5 shelf=0 slot=0 adapter=0 port=1 channel=0 13w6d: AAA/MEMORY: create_user (0x2036098) user='NULL' ruser='NULL' ds0=0 port='tty1' rem_addr='192.168.250.5' authen_type=ASCII service=LOGIN priv=1 initial_task_id='0', vrf= (id=0) 13w6d: AAA/AUTHEN/START (2523915680): port='tty1' list='' action=LOGIN service=LOGIN 13w6d: AAA/AUTHEN/START (2523915680): using "default" list 13w6d: AAA/AUTHEN/START (2523915680): Method=tacacs+ (tacacs+) 13w6d: TAC+: send AUTHEN/START packet ver=192 id=2523915680 13w6d: TAC+: ver=192 id=2523915680 received AUTHEN status = GETUSER 13w6d: AAA/AUTHEN (2523915680): status = GETUSER 13w6d: AAA/AUTHEN/CONT (2523915680): continue_login (user='(undef)') 13w6d: AAA/AUTHEN (2523915680): status = GETUSER When the username: AJ is entered, a GETPASS response is initiated. Username: aj Password: GETPASS message initiated on Site2-Core-SW1 13w6d: AAA/AUTHEN (2523915680): Method=tacacs+ (tacacs+) 13w6d: TAC+: send AUTHEN/CONT packet id=2523915680 13w6d: TAC+: ver=192 id=2523915680 received AUTHEN status = GETPASS 13w6d: AAA/AUTHEN (2523915680): status = GETPASS 13w6d: AAA/AUTHEN/CONT (2523915680): continue_login (user='aj') 13w6d: AAA/AUTHEN (2523915680): status = GETPASS When the user enters the password, the authentication is failed. Username: aj Password: 75 % Authentication failed. Fail authentication message on Site2-Core-SW1 13w6d: AAA/AUTHEN (2523915680): Method=tacacs+ (tacacs+) 13w6d: TAC+: send AUTHEN/CONT packet id=2523915680 13w6d: TAC+: ver=192 id=2523915680 received AUTHEN status = FAIL 13w6d: AAA/AUTHEN (2523915680): status = FAIL The following failed attempt logs in figure 3-17 are captured on the Site2 ACS server. It displaces the time the user AJ tried to connect to an unauthorized device in AJ_Site2 group. Figure 3-18 Failed Authentication Logs of User AJ From both the above test’s it proves that a user can only be able to established connectivity to network devices which he is authorized to. Hence the user AJ can only be authorized to devices in Site1 and not authorized to devices in Site2. 3.2.10 Round Trip Delay For Site1 Primary TACACS+ Server Perform a ping to the primary TACACS server from DS2 to observe the round-trip-delay. DS2#ping 192.168.250.70 Type escape sequence to abort. 76 Sending 5, 100-byte ICMP Echos to 192.168.250.70, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/1/2 ms DS2# The minimum round trip delay for Site1 devices for the primary TACACS Server is 1 ms. And the maximum round trip delay is 2 ms. 3.2.11 Round Trip Delay For Site1 Secondary TACACS+ Server Perform a ping to the secondary TACACS server from DS2 to observe the round-tripdelay. DS2#ping 192.168.250.132 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 192.168.250.132, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 4/4/8 ms DS2# The minimum round trip delay for Site1 devices for the secondary TACACS Server is 4 ms. And the maximum round trip delay is 8 ms. It can be concluded from the above tests that the maximum time taken by site1 network devices to communication to the primary AAA server within the same region over a secondary AAA server in different region is 4 times faster. Secondly, pointing network devices to primary AAA server within the same region encompasses the AAA communication messages within the region itself. Hence the network noise is limited to the region. 77 CHAPTER 4. PRE-PRODUCTION DEPLOYMENT AND NETWORK PERFORMANCE ANALYSIS This chapter describes the comparison test results and performance analysis of production devices communicating to a primary AAA server in a different region compared to preproduction deployment of AAA servers within a region. As enterprise networks are geographically distributed, the effect of round-trip authentication delay can be measured more significantly when devices from one region are pointing to a primary server in another region. To demonstrate this, two sites one in Penang and the other in Shanghai are considered as regional GAR sites for acceptance testing and three different regional devices from Penang, China and Bangalore are used as sub-regional GAR sites to demonstrate statistical analysis of data and performance improvement. The devices in GAR region were initially pointing to a primary server in a European region GER. Both these regions are integrated with WAN and ISP network. Hence to communicate a device from GAR to a AAA server in GER requires several network hops. For example, the following round-trip delay is recorded to reach a server in GER.A trace route is done from a device in Penang to the primary TACACS server in GER. A total of 12 hops were captured end to end with a minimum round trip delay of 220ms. Device-PG#traceroute 10.x.x.x Type escape sequence to abort. Tracing the route to aaa-ger-server.ger.corp.xyz.com (10..x.x.x) 1 device-v901.gar.xyz.com (10.x.x.x) 0 msec 0 msec 0 msec 2 core-v250.gar.xyz.com (132.x.x.x) 0 msec core-v251.gar.xyz.com (132.x.x.x) 4 msec 0 msec 78 3 devicea-tengig2-1.wan.xyz.com (10.x.x.x) 0 msec deviceb-tengig2-2.wan.xyz.com (10.x.x.x) 0 msec devicec-tengig2-1.wan.xyz.com (10.x.x.x) 4 msec 4 coreroutera-tengig2-5.wan.xyz.com (10.x.x.x) 0 msec 0 msec corerouterb-tengig2-6.wan.xyz.com (10.x.x.x) 0 msec 5 regcorerouter-tengig2-1.wan.xyz.com (10.x.x.x) 8 msec 4 msec 20 msec 6 regcorerouter-gigabitethernet0-0-0.wan.xyz.com (10.x.x.x) 4 msec 8 msec regcorerouter-gigabitethernet0-1-0.wan.xyz.com (10.x.x.x) 4 msec 7 corerouter-tunnel1109.wan.xyz.com (10.x.x.x) 228 msec corerouter-tunnel1109.wan.xyz.com (10.x.x.x) 224 msec corerouter-tunnel1109.wan.xyz.com (10.x.x.x) 212 msec 8 corerouter-gigabitethernet0-0-6.wan.xyz.com (10.x.x.x) 228 msec dc-routera-470.ger.xyz.com (143.x.x.x) 224 msec 228 msec 9 dc1-routera02-471.ger.xyz.com (143.x.x.x) 240 msec 232 msec 252 msec 10 dc1-corerouter.ger.xyz.com (10.x.x.x) 244 msec dc2-corerouter.ger.xyz.com (10.x.x.x) 220 msec dc3-corerouter.ger.xyz.com (143.x.x.x) 220 msec 11 dc3e-corerouter.ger.xyz.com (10.x.x.x) 228 msec 10.x.x.x 216 msec dc3e-corerouter2.ger.xyz.com (10.x.x.x) 228 msec 12 aaa-server.ger.corp.xyz.com (10.x.x.x) 216 msec 10.x.x.x 224 msec aaa-server.ger.corp.xyz.com (10.x.x.x) 216 msec Device-PG# A ping is executed to observe the minimum round trip network delay of 220 ms. Device-PG#ping 10.x.x.x Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 10.x.x.x, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 220/224/232 ms Device_PG#exit 4.1 Analysis Of Round-Trip-Delay At Different Time Periods For A AAA Server Outside Region: Three GAR sub-regional sites namely Penang, China and Bangalore are used to analyze the RTD to communicate a AAA server in a GER regional site. Three different devices from the same sub region were used to get an average Round-Trip-Delay. The RTD 79 analysis is done at four different period’s viz. morning hours, peak hours, afternoon hours and off hours. 4.1.1 Analysis Of RTD During Morning Hours The following table provides the stats captured during the morning hours. Table 4-1 Average Morning Hour Stats For A AAA Server Outside A Region Morning_Hours MIN MAX AVG Device1_PG 239 267 248 Device2_PG 237 265 244 Device3_PG 231 248 237 Avg_PG 236 260 243 Device1_BGA 249 299 264 Device2_BGA 235 255 247 Device3_BGA 233 384 285 Avg_BGA 239 313 265 Device1_CHN 223 324 255 Device2_CHN 222 270 236 Device3_CHN 224 249 233 Avg_CHN 223 281 241 Avg_GAR_Morning_RTD 233 285 250 The following graph displays the averages RTD for the stats captured during morning hours and a total average of 250ms for all GAR devices is analyzed to reach a server in an outside region. 80 Average RTD For All GAR Devices To Reach TACACS+ Server Outside the GAR Region During Morning Hours 330 320 310 300 Time 290 in MilliSeconds(ms)280 270 Avg_BGA, 265 260 Avg_GAR_Morning _RTD, 250 Avg_PG, 243 Avg_CHN, 241 250 240 230 220 MIN MAX AVG Figure 4-1 Average RTD During Morning Hours 4.1.2 Analysis Of RTD During Peak Hours The following table provides the stats captured during the peak hours. 81 Table 4-2 Average Peak Hour Stats For A AAA Server Outside A Region Peak_Hours Device1_PG Device2_PG Device3_PG Avg_PG Device1_BGA Device2_BGA Device3_BGA Avg_BGA Device1_CHN Device2_CHN Device3_CHN Avg_CHN Avg_GAR_Peak_RTD MIN MAX 228 228 227 228 235 237 235 236 224 235 239 233 232 AVG 231 240 274 248 680 657 716 684 256 340 241 279 404 232 235 242 236 374 433 463 423 243 265 240 249 303 And the following graph displays the averages RTD for the stats captured during peak hours and a total average of 250ms for all GAR devices is analyzed to reach the same server in the outside region. Average RTD For All GAR Devices To Reach TACACS+ Server Outside the GAR Region During Morning Hours 330 320 310 300 Time 290 in 280 MilliSeconds(ms) 270 260 250 240 230 220 Avg_BGA, 265 Avg_GAR_Morning _RTD, 250 Avg_PG, 243 Avg_CHN, 241 MIN MAX AVG Figure 4-2 Average RTD During Peak Hours 82 4.1.3 Analysis Of RTD During Afternoon Hours The following table provides the stats captured during the afternoon hours. Table 4-3 Average Afternoon Hour Stats For A AAA Server Outside A Region Afternoon_Hours MIN MAX AVG Device1_PG 212 220 214 Device2_PG 215 215 215 Device3_PG 214 217 215 Avg_PG 214 217 215 Device1_BGA 224 224 224 Device2_BGA 224 226 228 Device3_BGA 228 231 232 Avg_BGA 225 227 228 Device1_CHN 212 214 220 Device2_CHN 215 216 215 Device3_CHN 215 218 215 Avg_CHN 214 216 217 Avg_GAR_Afternoon_RTD 218 220 220 And the following graph displays the averages RTD for the stats captured during afternoon hours and a total average of 220ms for all GAR devices is analyzed to reach the same server in the outside region. Average RTD For All GAR Devices To Reach TACACS+ Server Outside the GAR Region During Afternoon Hours 230 Avg_BGA, 228 225 Time in 220 MilliSeconds(ms) Avg_GAR_Afterno on_RTD, 220 Avg_CHN, 217 Avg_PG, 215 215 210 MIN MAX AVG Figure 4-3 Average RTD During Afternoon Hours 83 4.1.4 Analysis Of RTD During Off Hours The following table provides the stats captured during the off peak hours. Table 4-4 Average Off Peak Stats For A AAA Server Outside A Region Off Peak Hours Device1_PG Device2_PG Device3_PG Avg_PG Device1_BGA Device2_BGA Device3_BGA Avg_BGA Device1_CHN Device2_CHN Device3_CHN Avg_CHN Avg_GAR_Off Peak_RTD MIN MAX 156 156 155 156 177 178 153 169 127 128 128 128 151 AVG 160 157 224 180 178 203 232 204 182 128 130 147 177 157 156 173 162 177 184 174 178 140 128 129 132 158 And the following graph displays the averages RTD for the stats captured during off peak hours and a total average of 158ms for all GAR devices is analyzed to reach the same server in the outside region. Average RTD For All GAR Devices To Reach TACACS+ Server Outside the GAR Region During Off Hours 220 195 Avg_BGA, 178 Time 170 in MilliSeconds(MS) 145 Avg_PG, 162 Avg_GAR_Off Peak_RTD, 158 Avg_CHN, 132 120 MIN MAX AVG Figure 4-4 Average RTD During Off Hours 84 4.1.5 Analysis Of Average RTD During Different Time Periods The following table provides the average stats of the different time period analysis performed above. Table 4-5 Average Stats For A AAA Server Outside A Region Average RTD Avg_Morning_RTD Avg_Peak_RTD Avg_Afternoon_RTD Avg_OffHours_RTD MIN MAX 233 232 218 151 AVG 285 404 220 177 250 303 220 158 A graph representation is show below displaying the averages of all time periods. Average RTD For All GAR Devices To Reach TACACS+ Server Outside the GAR Region At Different Time Periods 425 375 325 Avg_Peak_RTD, 303 Avg_Morning_RTD , 250 Avg_Afternoon_RT D, 220 Avg_OffHours_RT D, 158 Time 275 in MilliSeconds(MS) 225 175 125 MIN MAX AVG Figure 4-5 Average RTD At Different Time Periods It is observed that the Average RTD during the peak time and the morning time is almost the same which is greater than 250ms. One of the reasons is that even though the networks in both GAR and GER regions are geographically distributed they are busy during that period. It can also be observed that the RTD is low during afternoon and 85 during off hours. Finally, it can be observed that the average round trip delay is above 219ms while trying to reach a AAA server configured in different region during afternoon hours or above 250ms when trying to reach a server during peak hours. 4.2 Analysis Of Hop Count At Different Time Periods For A AAA Server Outside Region: Trace route was performed from the sub-regional GAR devices in Penang, China and Bangalore to the AAA server in the outside region. The trace was performed at different time periods during morning, afternoon, peak and off hours. Table 4-6 Average Hop Count For A AAA Server Outside A Region Avg_Hop_Count Device_PG Device_BGA Device_CHN Morning Afternoon Peak 12 12 12 9 9 9 13 13 13 Off Hours 12 9 13 Periodic Hops Count To A TACACS+ Server Outside GAR Region 14 Device_CHN, 13 Device_PG, 12 12 10 Device_BGA, 9 No Of Hops 8 6 4 2 0 Morning Afternoon Peak Off Hours Figure 4-6 Periodic Representation Of Hop Count From Different Regions 86 From the above graph, it was observed that the number of hops were constant from all the devices in different region. The lowest number of hops was 9 observed from devices in Bangalore and the highest no of hops were observed from China with a hop count of 13. 4.3 Deploying Regional Pre-Production AAA Servers Two AAA servers are integrated in Penang and Shanghai as GAR regional servers. All devices in China are pointed to the Shanghai AAA server as primary server and configured with backup server in Penang. Similarly, the rest of the GAR regional devices are pointed to the Penang server as the primary server and Shanghai server as the secondary. As an example, the following round-trip delay is recorded to reach the server within the region. A trace route is done from a device in Penang and a total of 4 hops were captured with a minimum round trip delay of maximum 4ms. Device-PG1#traceroute 198.x.x.x Type escape sequence to abort. Tracing the route to aaa-server.pg.corp.xyz.com (198.x.x.x) 1 dist-sw-v901.pg.corp.xyz.com (10.x.x.x) 0 msec 0 msec 4 msec 2 core-router1- v251.pg.corp.xyz.com (132.x.x.x) 0 msec 0 msec core-router2-v250.pg.corp.xyz.com (132.x.x.x) 0 msec 3 rtap-tengig2-2.wan.pg.corp.com (10.x.x.x) 0 msec rtbp-tengig2-2.wan.pg.corp.com (10.x.x.x) 0 msec 0 msec 4 aaa-server.pg.corp.xyz.com (198.x.x.x) 0 msec 0 msec 0 msec Device-PG1# A ping is performed to get the round trip delay to the AAA server with a maximum RTD of 4ms. Device-PG#ping 198.x.x.x 87 Type escape sequence to abort. Sending 5, 100-byte ICMP Echos to 198.x.x.x, timeout is 2 seconds: !!!!! Success rate is 100 percent (5/5), round-trip min/avg/max = 1/1/4 ms Device-PG#exit 4.4 Analysis Of Round-Trip-Delay At Different Time Periods For A AAA Server Within A Region: The same devices from the three sub-regions Penang, China and Bangalore were used to analyze the RTD to communicate the AAA server integrated in the GAR region. The Round-Trip-Delay was captured at four different time periods viz. morning hours, peak hours, afternoon hours and off hours. 4.4.1 Analysis Of RTD During Morning Hours The following table provides the stats captured during the morning hours. Table 4-7 Average Morning Hour Stats For A AAA Server Within A Region Morning_Hours MIN Device1_PG Device2_PG Device3_PG Avg_PG Device1_BGA Device2_BGA Device3_BGA Avg_BGA Device1_CHN Device2_CHN Device3_CHN Avg_CHN Avg_GAR_Morning_RTD MAX 3 4 3 9 27 28 28 28 26 28 28 27 21 AVG 13 4 4 7 37 29 28 31 27 32 30 30 23 5 4 3 4 29 28 28 28 26 29 28 28 20 88 The following graph displays the averages RTD for the stats captured during morning hours and a total average of 20ms for all GAR devices is analyzed to reach a server within the same region. RTD To Reach TACACS+ Server In The Same Region During Morning Hours 35 30 Avg_BGA, 28 Avg_CHN, 28 25 Time in MilliSeconds(MS) Avg_GAR_Mornin g_RTD, 20 20 15 10 5 Avg_PG, 4 0 MIN MAX AVG Figure 4-7 Average RTD During Morning Hours 4.4.2 Analysis Of RTD During Peak Hours The following table provides the stats captured during the peak hours. 89 Table 4-8 Average Peak Hour Stats For A AAA Server Within A Region Peak_Hours Device1_PG Device2_PG Device3_PG Avg_PG Device1_BGA Device2_BGA Device3_BGA Avg_BGA Device1_CHN Device2_CHN Device3_CHN Avg_CHN Avg_GAR_Peak_RTD MIN MAX 3 3 4 3 27 27 29 28 27 28 28 28 20 AVG 44 48 52 48 115 283 104 167 41 29 34 35 83 13 15 16 15 76 91 48 72 30 28 29 29 39 The following graph displays the averages RTD for the stats captured during peak hours and a total average of 40ms for all GAR devices is analyzed to reach a server within the same region. RTD To Reach TACACS+ Server In The Same Region During Peak Hours 180 160 140 120 Time in 100 MilliSeconds(MS) 80 Avg_BGA, 72 60 Avg_GAR_Peak_R TD, 39 Avg_CHN, 29 Avg_PG, 15 40 20 0 MIN MAX AVG Figure 4-8 Average RTD During Peak Hours 90 4.4.3 Analysis Of RTD During Afternoon Hours The following table provides the stats captured during the afternoon hours. Table 4-9 Average Afternoon Hour Stats For A AAA Server Within A Region MIN Afternoon_Hours Device1_PG Device2_PG Device3_PG Avg_PG Device1_BGA Device2_BGA Device3_BGA Avg_BGA Device1_CHN Device2_CHN Device3_CHN Avg_CHN Avg_GAR_Afternoon_RTD MAX AVG 3 4 3 4 3 4 5 4 4 3 27 26 27 27 23 24 24 24 18 4 28 28 30 29 24 24 24 24 19 4 27 27 28 27 23 28 24 25 19 The following graph displays the averages RTD for the stats captured during afternoon hours and a total average of 19ms for all GAR devices is analyzed to reach a server within the same region. 91 RTD To Reach TACACS+ Server In The Same Region During Afternoon Hours 35 30 Avg_BGA, 27 25 Avg_CHN, 25 Time 20 in MilliSeconds(ms) 15 Avg_GAR_Afterno on_RTD, 19 10 5 Avg_PG, 4 0 MIN MAX AVG Figure 4-9 Average RTD During Afternoon Hours 4.4.4 Analysis Of RTD During Off-Peak Hours The following table provides the stats captured during the off peak hours. Table 4-10 Average Off Peak Stats For A AAA Server Within A Region Off Peak Device1_PG Device2_PG Device3_PG Avg_PG Device1_BGA Device2_BGA Device3_BGA Avg_BGA Device1_CHN Device2_CHN Device3_CHN Avg_CHN Avg_GAR_Off Peak_RTD MIN MAX 1 1 1 1 25 25 24 25 23 24 24 24 17 AVG 1 1 4 2 28 27 28 28 23 32 24 26 19 4 1 2 2 26 26 24 25 23 25 24 24 17 92 The following graph displays the averages RTD for the stats captured during off peak hours and a total average of 17ms for all GAR devices is analyzed to reach a server within the same region. RTD To Reach TACACS+ Server In The Same Region During Off Peak Hours 30 Avg_BGA, 25 25 Avg_CHN, 24 20 Avg_GAR_Off Peak_RTD, 17 Time 15 in MilliSeconds(MS) 10 5 Avg_PG, 2 0 MIN MAX AVG Figure 4-10 Average RTD During Off-Peak Hours 4.4.5 Analysis Of Average RTD During Different Time Periods The following table provides the average stats of the different time period analysis performed above. Table 4-11 Average Stats For a AAA Server Within A Region Average RTD Avg_Morning_RTD Avg_Peak_RTD Avg_Offpeak_RTD Avg_ Afternoon_RTD MIN 19 19 16 18 MAX 26 67 17 19 AVG 20 40 18 18 93 A graph representation is show below displaying the averages of all time periods. Average RTD For All GAR Devices To Reach TACACS+ Server In The Same Region 90 80 70 60 Time 50 in MilliSeconds(MS) 40 Avg_Peak_RTD, 39 30 Avg_Morning_RTD, 20 20 Avg_Afternoon_RT D, 19 Avg_Off Peak_RTD, 17 10 0 MIN MAX AVG Figure 4-11 Average RTD At Different Time Periods It is observed that the Average RTD during the off-peak time and the afternoon time is 18ms which is the lowest RTD. One of the main reasons is that the devices are within the same geographical location. Moreover, it can be observed that the average round trip delay is above 40ms while trying to reach AAA server configured in the same region during peak hours. 94 4.5 Analysis Of Hop Count At Different Time Periods For A AAA Server In The Same Region: Trace route was performed from the sub-regional GAR devices in Penang, China and Bangalore to the AAA server integrated in the same GAR region. The trace was performed at different time periods during morning, afternoon, peak and off-peak hours. Table 4-12 Average Hop Count For A AAA Server Within A Region Avg_Hop_Count Morning Afternoon Peak 4 6 5 Device_PG Device_BGA Device_CHN 4 6 5 Off Hours 4 4 6 6 5 5 Periodic Hops Count To A TACACS+ Server In The Same Region 14 Device_CHN, 13 Device_PG, 12 12 10 Device_BGA, 9 No Of Hops 8 6 4 2 0 Morning Afternoon Peak Off Hours Figure 4-12 Periodic Hop Count To A TACACS+ Server In the Same Region From the above analysis, it was observed that the number of hops were constant from all the devices in different sub-region to reach a AAA server within the same region. The 95 lowest number of hops was 4 observed from devices in Penang where the AAA server is actually integrated and the highest no of hops were observed from Bangalore with a hop count of 6. 4.6 Improvement In Network Performance The improvement in network performance is measured in both end-to-end Round-TripDelay and Hop Count. Firstly, the network performance gain is measured by the decrease in average RTD of having a AAA server outside the region compared to having a server within a region. The following table shows the summary of the Average RTD at different time periods for devices in GAR (Penang, China and Bangalore), communicating to a AAA server outside the region and to a server inside the region. Table 4-13 Average Performance Improvement in RTD Average Performance Improvement Avg_Morning_RTD Avg_Peak_RTD Avg_Afternoon_RTD Avg_OffHours_RTD Outside Region 250 303 220 158 Inside Region 20 39 19 17 The following graph shows statistical representation of the performance improvement in RTD. 96 Average Performance Improvement in RTD 350 300 250 Time 200 in MilliSeconds(MS) 150 100 50 0 Outside Inside Region Avg_Morning_RTD 250 20 Avg_Peak_RTD 303 39 Avg_Afternoon_RTD 220 19 Avg_OffHours_RTD 158 17 Figure 4-13 Average Performance Improvement in RTD It can be observed that a significant drop in average RTD is achieved. The Average RTP of 303ms at peak time is reduced to 39ms by having a dedicated server for the GAR devices. Similarly, the Average RTD of 158ms during Off peak time is dropped to 17ms. It can be concluded that the average performance gain during peak time is improved by 87.1% and the average performance gain during off peak time is improved by 89.2%. Secondly, the performance improvement is also measured by the number of hops reduced from reaching a server outside the region compared to reaching a server inside the same region. 97 The following table shows the summary of the Average Hop Count at different time periods for devices in GAR (Penang, China and Bangalore), communicating to a AAA server outside the region and to a server inside the region. Table 4-14 Average Hop Count Improvement Avg_Hop_Count Device_PG Device_BGA Device_CHN Outside Region Inside Region 12 9 13 4 5 6 The following graph shows statistical representation of the performance improvement in Hop Count. Average Hop Count Improvement 14 12 10 8 Hop Count 6 4 2 0 Outside Region Inside Region Device_PG 12 4 Device_BGA 9 5 Device_CHN 13 6 Figure 4-14 Average Hop Count Improvement It can be observed that a significant drop in average hop count is achieved. The Average hop count of 12 for devices in Penang communicating with server outside the region is 98 reduced to 4 by communicating to server within the same region. Similarly, the devices in Bangalore and China sub-regional sites are reduce to 5 and 6 hops respectively. It can be concluded that the average hop count for devices in PG, where the AAA server is actually integrated, is improved by 75%. And the average hop count for devices in sub-regions, Bangalore and China, is improved by ~50%. 99 CHAPTER 5. CONCLUSION TACACS+ provides an additional layer of security to network devices by enabling AAA services to network users. It allows access to network services to be regulated on a more granular basis. Network users connecting to any network device can be authenticated and authorized against the user database with a custom policy applied. This policy not just controls what devices a user is authorized to access but also controls what services on a particular device that a user can access. These administrative policies can be managed by deploying a centralized TACACS+ architecture with a dedicated master server acting as a centralized point for global administration and replicating the database to dedicated region primary and secondary slave servers for authentication, authorization and accounting. By deploying these regional slave servers it is analyzed that average performance gain during peak time is improved by 87.1% and the average performance gain during offpeak time is improved by 89.2%. Secondly, the average hop count for devices in Penang, where the AAA server is actually integrated, is improved by 75%. And the average hop count for devices in sub-regions, Bangalore and China, is improved by ~50%. Finally, by deploying a hierarchal network design, the TACACS+ back and forth request and response AAA messages encompasses with a dedicated region. Hence the network noise is limited to a region there by reducing immense noise on core layer. The analysis done in this project can be extended by virtualize these AAA server and analyzing the stress testing to provide an additional layer of security and thereby reducing hardware cost. 100 ACRONMYS Acronym Definition AAA ACL ACP ACS AS BGP CSV DS EIGRP FTP GAR GER HSRP IOS IP LAN MTU NAD NDG NMS PPP RADIUS RTD SNMP TACACS TCP VLAN VTP WAN Authentication, Authorization and Accounting Access Control List Access Control Protocol Access Control System Autonomous System Border Gateway Protocol Comma Separated Value Distribution Switch Enhanced-Interior Gateway Routing Protocol File Transfer Protocol Greater Asian Region Greater European Region Host Standby Routing Protocol Internet Operating System Internet Protocol Local Area Network Maximum Transmission Unit Network Access Device Network Device Group Network Management System Point to Point Protocol Remote Access Dial in User Services Round Trip Delay Simple Network Management Protocol Terminal Access Controller Access Control System Transmission Control Protocol Virtual LAN Virtual Trunking Protocol Wide Area Network 101 REFERENCES [1] User Guide for Cisco Secure ACS for Windows 4.0 http://www.cisco.com/en/US/docs/net_mgmt/cisco_secure_access_control_serve r_for_windows/4.0/user/guide/o.html#wp82226 [2] Building Cisco Multilayer Switched Networks, by Richard Froom, Balaji Sivasubramanian and Erum Frahim (isbn: 1587051508) [3] Network Design and Case Studies ciscopress.com (isbn: 1578701678) [4] Enhanced Interior Gateway Routing Protocol http://www.cisco.com/en/US/products/ps6630/products_ios_protocol_option_ho me.html [5] Deploying IGRP/E-IGRP- Session 2208 http://www.cisco.com/en/US/prod/collateral/iosswrel/ps6537/ps6554/ps6599/ps6 630/prod_presentation0900aecd80310f03.pdf [6] Cisco Secure ACS 4.2 – Installation Guide http://www.cisco.com/en/US/docs/net_mgmt/cisco_secure_access_control_serve r_for_windows/4.2/installation/guide/windows/install.html [7] TACACS & RADIUS Comparisons http://www.cisco.com/en/US/tech/tk59/technologies_tech_note09186a0080094e9 9.shtml [8] TACACS+ Concepts http://www.xperiencetech.com/online/tacacs.htm
© Copyright 2026 Paperzz