Modular Network Design Campus Network Design Modules are analogous to building blocks of different shapes and sizes; when creating a building, each block has different functions Designing one of these blocks is a much easier task than designing the entire building Each block might be used in multiple places, saving time and effort in the overall design and building process The blocks have standard interfaces to each other so that they fit together easily If the requirements for a block change, only that block needs to change other blocks are not affected. Similarly, a specific block can be removed or added without affecting other blocks. 2003, Cisco Systems, Inc. All rights reserved. 2-1 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-2 Modular Network Design Access Layer Entry point to the network Shared bandwidth includes port security, DHCP snooping, Dynamic ARP inspection, IP source guard. Layer 2 services Filtering VLAN membership Access layer Provides user and workgroup access to the resources of the network Distribution layer Implements the organization's policies, and provides connections between workgroups and between the workgroups and the core Core layer Provides high-speed transport between distribution-layer devices and to core resources 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-3 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-4
Distribution Layer The Core Layer High availability, fast path recovery, load balancing, QoS, and security Route summarization and packet manipulation Redistribution point between routing domains Packet filtering and policy routing to implement policy-based connectivity Terminate VLANs Aggregates distribution layer switches. Implements scalable protocols and technologies and load balancing. High-speed layer 3 switching using 10-Gigabit Ethernet. Uses redundant L3 links. 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-5 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-6 Enterprise Campus Infrastructure Server Farm Distribution Layer Deploy high to mid-range switches. Make switching and links entirely redundant. 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-7 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-8
Server Farm Access Layer Small Campus Network <200 end devices Collapsed core Catalyst 3560 and 2960G switches for access layer Cisco 1900 and 2900 routers to interconnect branch/wan Deploy midrange switches. Dual home all servers. 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-9 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-10 Example Small Enterprise Medium Campus Backbone 200-1000 end devices 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-11 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-12
Example Medium Enterprise Large Campus Design 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-13 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-14 Large-Scale Layer 3 Switched Campus Backbone Example Large Enterprise 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-15 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-16
Data Center Infrastructure A Collapsed Core Core layer high-speed packet switching backplane Aggregation layer service module integration, default gateway redundancy, security, load balancing, content switching, firewall, SSL offload, intrusion detection, network analysis Access layer connects servers to network Access Layer Switch Block 1 Access Layer Switch Block 2 Core Connectivity Distribution/Core Layer Distribution/Core Layer Core Connectivity 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-17 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-18 A Collapsed Core (Cont.) The Dual Core Switch Block 1 Switch Block 2 Switch Block 1 Switch Block 2 Access Layer Access Layer Core Connectivity Distribution/Core Layer Distribution/Core Layer Core Connectivity 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-19 Core Block Subnet A Subnet B 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-20
The Dual Core (cont.) Task List Switch Block 1 Switch Block 2 Core Block Subnet A Subnet B 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-21 Requirement Preliminary Design Decision New Physical Diagram New Logical Diagram Device List / Price VLAN / IP Address Spanning Tree Router Redundancy 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-22 Design Objective Business Requirement Why do you want to build a network? Too often people build networks based on technological, rather than business consideration Requirement Money : one of most important design constraints on any network. Upper limit to what can be accomplished. Geography : Where are the users? Where are the services they want to access? How are the users organized geographically? Design Objective Requirement Installed Base : What technology exists today? Why does it need to be changed? How much of the existing infrastructure must remain? There are geographical constrains, such as the location and accessibility of the computer rooms and LAN rooms Bandwidth : Get as much information as possible about all of the major traffic patterns and how much volume they involve. (average rates at the peak periods of the day. Security : what are the security requirements? 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-23 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-24
Design Philosophy Elements of Reliability Scalable Where switches for different workgroups are interconnected. Hierarchical designs Manageable You want to see what is going on throughout the network easily. Demand simple, rational address schemes. Reliable Eliminate key single points of failure. Defining Reliability Fault tolerance Performance and capacity Reliability Issue Failure How frequency the network fails to meet the business requirement, and how badly it fails. Failure that are very short duration, but which interrupt key application for much longer periods. If a key application is sensitive to latency, then a slow network will be considered unreliable if it never breaks. 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-25 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-26 Elements of Reliability Design Principles Reliability Issue Performance Two importance factors : bandwidth and latency Bandwidth is the amount of data that the network can transmit per unit time. Latency is the length of time it takes to send that data from end to end. Business requirement and not the technology determine what is the best way. Conclusion : Latency, Bandwidth, length of outage Technique for improving reliability is to duplicate key pieces of equipement. Plan phase The detailed network requirements are identified, and the existing network is reviewed. Design phase The network is designed according to the initial requirements and additional data gathered during analysis of the existing network. The design is refined with the client. Implement phase The network is built according to the approved design. Operate phase The network is operational and is being monitored. This phase is the ultimate test of the design. Optimize phase During this phase, issues are detected and corrected, either before problems arise or, if no problems are found, after a failure has occurred. Redesign might be required if too many problems exist. Retirement phase Although not part of the PDIOO acronym, this phase is necessary when part of the network is outdated or is no longer required. 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-27 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-28
Design Principles Task in Network Design 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-29 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-30 Determine Requirement Determine Requirement Applications that are to run on the network Internet connections required Addressing restrictions, for example, the use of private Internet Protocol (IP) version 4 (IPv4) addresses Support for IP version 6 (IPv6) addresses Other protocols that are to run on the network (for example, routing protocols) Cabling requirements Redundancy requirements Use of proprietary equipment and protocols Existing equipment that must be supported Network services required, including quality of service (QoS) and wireless How security is to be integrated into the network Network solutions required (for example, voice traffic, content networking, and storage networking) Support for existing applications while new ones are being phased in Bandwidth availability Budget Capital (for new equipment) and operating (for ongoing expenses). Schedule This could include the phasing out of older applications, hiring of new personnel, and so forth. People Considerations include who will install and operate the network, what skills they have, whether they require training, whether any of these tasks will be outsourced, and so forth. Legal Issues include any restrictions on the use and storage of data collected, whether the organization has contractual obligations or opportunities related to the network (for example, long-term maintenance or lease contracts), and so forth. History Factors include examining the existing network's structure and determining whether any person or group will block changes or additions. Policies Consider whether current organizational policies might restrict the network design. 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-31 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-32
Analyzing the Existing Network Preparing the Preliminary Design If this is a redesign of an existing network, the current network must be analyzed and understood. An existing network is likely to restrict the network design Should analyze the network to determine both what is good and what should be changed. Examining documentation about the existing network and discussing it with users, administration staff, and other stakeholders is important Do an audit of the network. The audit can identify items such as the protocols that are running Preliminary design involves considering all the network requirements and constraints (including the budget), and determining viable alternative solutions. The network owner is then consulted, and together an optimal solution is chosen; this solution is later developed into the final design. Two models that can be used for network design are examined A top-down approach to network design means that requirements are considered first, with the applications and network solutions that will run on the network driving the design. A bottom-up approach would first select devices, features, cabling, and so on, and then try to fit the applications onto this network. A bottom-up approach can lead to redesign if the applications are not accommodated properly. This approach can also result in increased costs by including features or devices that are not required and would have been excluded had the network requirements analysis been completed. 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-33 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-34 Completing the Final Design Development Completing the Final Design Development producing detailed drawings, configuration specifications, costing, addressing plans, and any other information required for implementation. Physical Network diagram Network redundancy Physical connectivity to existing network Post-design hardware inventory documenting device location, type, number of ports and type of ports Logical Addressing scheme(s) Supported Protocol(s) Routing protocol(s) Virtual Local Area Network (VLAN) architecture Design Guide Security requirements Redundancy requirements and design Logical connectivity within the Local Area Network (LAN) environment Implementation of any required equipment software advanced features Connectivity to the gateway Configuration of each device 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-35 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-36
Deploying the Network Monitoring and Redesigning The deployment plan must include details of what is to be done and how it is to be done. For example, if new cabling is required, the procedure to run the cable and the location where it is needed must be fully documented. Scheduling is important, not only to identify when things will be done but also to determine who will do them, and what impact the deployment will have on the existing network. After the network is operating, baseline operational statistics should be gathered The network should then be monitored for anomalies and problems. If problems that require redesign occur, or if requirements change or are added, the appropriate design changes must be made and the entire design process should be repeated for that portion of the network. 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-37 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-38 Maintaining Design Documentation The design should be documented throughout the process. Documentation should include the following items: All the agreed-to requirements and constraints The state of the existing network, if any Preliminary design options and a brief review of why the final design was chosen Final design details Results of any pilot or prototype testing Deployment plans, schedules, and other implementation details Monitoring requirements Any other pertinent information 2003, Cisco Systems, Inc. All rights reserved. BCMSN v2.0 2-39