Now that we have reviewed the basics of the networking protocols that would be fundamental to build an Enterprise network, let's discuss the considerations for building an IP network.

We see so many networks around us. Each network has a specific purpose for which it is built. For example, the primary purpose of the networks that we see in computer labs is to provide access to shared resources, most notably printers and data storage. The networks in the manufacturing plant are meant to carry control signals for the various plant machinery that are connected on the network. The military and defense networks have a totally different purpose.

Since the networks are supposed to deliver different services to the end users, the design of the network will be different, and will be defined by the characteristics of the services to a large extent. Hence, the starting point for planning a network is to define the services that the network will offer, so that the network can be built accordingly.

Once the network is built, and starts offering services to the end users, it needs to be operated, and changes need to be made on the network on a day-to-day basis. The operations include monitoring the network for critical network parameters, and taking corrective action in case of network incidents such as outages/performance degradations. The changes might also include adding new services or deleting or modifying any existing services on the network. The network operations depend upon the way the network is designed, and the services it is running. For example, for a network that is not built with adequate redundancy, the operations approach has to be very different than that for a network that has enough resiliencies built in the network.

These concepts have been widely described and used in the frameworks used for network architectures, for example, for the Services, Network, Operations (SNO) approach, or the Services Oriented Network Architecture (SONA) framework proposed by Cisco.

As discussed in the previous section, the network is built for a specific purpose. Operating the network involves making changes to the network parameters, and sometimes design, to meet new business/application requirements, and finally the network is either replaced by a new design or incorporates a new technology. Since the network is dynamic, it is important to have a systematic approach based on the different phases of the network. Different approaches have been proposed by different vendors, but almost all of them are essentially overlapping and similar. In this section, we will cover Cisco's PPDIOO approach for network lifecycle, as it is the most comprehensive approach, and is a superset of other approaches within the scope of the network lifecycle. PPDIOO is an acronym for Prepare, Plan, Design, Implement, Operate, and Optimize.

Other forms of the lifecycle approach that are simplified versions of the PPDIOO approach are the Plan, Build, Manage (PBM) approach where some of the stages of the PPDIOO approach are combined into the three phases of the PBM approach.

The network lifecycle approach provides several key benefits aside from keeping the design process organized. Some of the benefits of using a structured lifecycle approach to network design are as follows:

• Lowering the total cost of network ownership: Businesses have always used the total cost of ownership (TCO) approach to take decisions. With IT becoming more and more relevant to business today, their IT decisions have to follow the approach of TCO rather than a pure capital expenditure (CAPEX) approach. This means that the operational expenses (OPEX) that have to be incurred while running and maintaining the network are also an important factor in the overall network approach. The network lifecycle approach helps in lowering the TCO by:
  • Identifying, evaluating, and choosing the right technology options.
  • Developing and documenting a design that is aligned with the business/service requirements.
  • Building implementation plans that can minimize the risk of implementation, thus avoiding cost and time overruns.
  • Planning for the operations as an integral part of network design so as to improve the operational efficiency by choosing the right set of tools, and operational skills required.
• Increasing services uptime: Downtime or outages are the most dreaded terms in network operations, as they causes service disruption resulting in loss of revenue and goodwill. A network lifecycle approach can help reduce downtime by:
  • Identifying the network elements that need to be highly available for service availability, and designing the network for redundancy of such elements.
  • Planning the operational skills required for the network, and ensuring that the Network Operations Center (NOC) staff has the right skills.
• Improving business agility: As businesses are faced with dynamic market trends, IT needs to be able to support business quickly and efficiently. This agility means that the network should have the ability to make changes to the way existing services are delivered on the network, or the ability to quickly add new services to the network based on the business requirements. A lifecycle approach helps provide this agility to the network by:
  • Capturing the business and technology requirements and their dependencies.
  • Developing detailed designs for each service at a block level and at a configuration level such that new services can be added without impacting the existing services.
  • Defining the horizontal and vertical scaling options for the network design during the design phase itself so that capacity can be quickly added to the network when required by the applications.
  • Creating operational run books and bringing in operational efficiencies through the proper use of tools and the right resource skills.

The PPDIOO approach consists of six phases as depicted in Figure 14, which are described as shown in the following diagram:

It is said with reference to the OSI layers that there is an eighth layer that rides preceding the application layer, and that is the business layer, as this is the layer that will define what applications are to be used on the network. The prepare phase tries to capture the business layer and the technological requirements of the underlying network infrastructure.

The prepare and the plan phases of the network lifecycle talk about the future and then hand it over to the next phases, which are concerned with how to build the present network so that it can meet the future requirements.

The prepare phase involves establishing the organizational requirements from a business perspective, and developing an appropriate technology strategy. The following are some examples of questions to be answered in this stage:

• What is the vision of the company?
• What are the business goals of the company today, and anticipating the goals and IT requirements in the future?
• What is the cloud strategy for the organization?
• Would the organization want to own the network assets and build a data center, or just host the applications in an outsourced data center?
• What will be the model of the DC outsourcing? Infrastructure as a Service or Platform, or Software as a Service?
• What is the communications strategy of the company? Would the company want to move to cloud-based models for its internal communications?
• What will be the WAN strategy for the network? Would the WAN links be owned or on a shared network?
• What is the Operations strategy for the organization?

The end goal of this stage is to develop a network strategy by comparing the different options and to propose an architectural view that identifies the various technologies to be used in the network, and the interdependencies between the various technologies. This phase also covers a lot of financial analysis, and building business cases for the decisions as all decisions have to be backed by sound financial reasoning.

By preparing for the network rollout in this manner, the company has a fair view of the budgetary requirements for the project in terms of time, money, and resources, and a long-term roadmap that can be leveraged as the business requirements change with time.

Most of these decisions are taken by the senior management and have already been taken by the time a network is being designed and implemented. Hence, we will not delve any further into these aspects in this book, but focus on the implications of the various technologies and how they impact the operational and business models in subsequent chapters.

This is the phase where the job of an actual network architect starts. This phase involves getting the right stakeholders together and documenting the network requirements with respect to the network goals, services, user needs, and so on. The plan phase involves identifying the sites, classifying them, and evaluating the existing network infrastructure if any to understand if the existing assets can be reused and redesigned for the new network. This phase also involves finalizing the hardware requirements for the network infrastructure devices.

Some of the questions that need to be answered at this stage are as follows:

• Who are the users of this network and what is the level of segmentation required between the various users?
• What are the services required by each group of users?
• Where are the users located?
• What is the hardware required to meet the user requirements?
• Where will the new hardware be installed?
• What are the power and space requirements at the locations?
• What are the existing services/network if any that need to be integrated/replaced by the new network?
• Where would the network be connected to the internet?
• What is the current security state of the company?
• What operational skills will be required to design/implement and operate the network?

Two important documents that are created during this phase are the Customer Requirement Document (CRD) that contains the detailed technical specifications of the network to be built, and the Site Requirement Specification (SRS) document that contains the physical, electrical, and environmental specifications for each site, where the equipment will be deployed. Site audits are done based upon the SRS documents to ensure that the sites are ready for the equipment to be installed, and any gaps/corrective action required is identified.

We will cover some of these topics in Chapter 2, Networks for Digital Enterprises, where we will describe the network requirements for a modern enterprise.

The business requirements have been drilled down into technical requirements until the planning phase. Now is the time to convert the technical requirements into the actual protocol-level details that will ensure that the network delivers the technical, functional, and performance requirements that the network is being designed for. In this phase of the network lifecycle, some of the most technical decisions are made such as:

• What should be the physical topology of the network?
• What should be the logical topology of the network?
• How should we plan for redundancy at the node level, site level, and at a service level?
• What should be the IP addressing schema for the network?
• What protocols should run on the network?
• How do we prioritize the different types of applications on the network?
• How do we segment the users on the network?
• How do we ensure security of the network devices?
• What management protocols should be run on the network?
• How would the different services be deployed on the network?
• How would we ensure that adding a new service does not impact any existing service?

It is in the design phase of the network lifecycle that the documents called the High-level design (HLD) and Low-level design (LLD) documents are made. The high-level design talks about the network design at a protocol level, and the low-level design talks about how to implement the design on the network devices and arriving at configuration templates. These design documents detail the design to meet the requirements of availability, reliability, flexibility, security, scalability, and performance.

The detailed design can also help in chalking out the day-to-day operational activities and network management processes, thereby simplifying network operations and helping to reduce OPEX and TCO. The design phase is also the phase when the design is validated on a staged network in the lab and configuration templates are fine-tuned.

Another important activity in the design phase is to define the test cases that will be executed on the network to ensure that the network is built as designed. The test case document is generally called an Acceptance Test Plan (ATP) document or a Network Ready for Use (NRFU) test plan. Having a documented test plan down to the details on how to execute the tests and what commands to run to validate the network implementation is crucial to ensure that the network will run as per the required specifications. A typical NRFU will have two parts: one covering the test cases that can be carried out on a standalone basis at each site, and the other part covering the end-to-end service testing across the entire network. The NRFU document can also add additional parts specific to network integration/service migration if the new network has to be integrated with any existing assets, or any existing services need to be migrated on the network that is being built.

We will cover these activities like choosing the right protocols and building the configuration templates based on these protocol choices in Chapters 3 to 9.

In the implement phase, the goal is to integrate devices and new capabilities in accordance with the design and without compromising network availability or performance. The implementation phase is where the actual implementation of the design starts. This phase includes deploying the network equipment and configuring it.

Site audits are reviewed and the actual implementation of the devices, including rack and stack power on testing is done in this stage. Some of the documents required for this stage include detailed installation documents for each type of equipment and the test process for each device type. Further, Network Implementation Plan (NIP) documents are created that are detailed documents for each site that is a part of the network. These documents lay down the list of equipment to be deployed at the site, the rack layouts, port connectivity diagrams, and the actual device configurations, along with the IP addressing and other variable parameters for each device at the site. The configurations are derived based on the configuration templates that were created and tested in the design phase. This document becomes the reference document for the implementation engineer who has to simply download the configuration onto the new devices and conduct the tests that are specific to the site.

Once all sites are up and ready, and the WAN connectivity is established, end-to-end service testing is conducted based on the NRFU test cases.

Any migration of existing services or any integration of networks with the existing network infrastructure is also carried out in this phase and the success validated against the test cases as defined in the NRFU document.

The operation phase of the network is when the actual users start using the network and the operations staff starts monitoring the network and services delivered on the network. It is important to have a multifaceted approach to network operations, that would include the domains of people, processes, and tools. The primary goal in the network operations phase is to maintain network and service uptime, at minimal cost. This can be done only if the organization has the right skills in the resources tasked with network operations, a structured process for the day-to-day activities so that the tasks are not dependent on individuals, but every single person carries a job in the same way as the others would. The tools aspect is essential to improve efficiencies, as mundane and routine tasks can be automated, thereby allowing the NOC resources to focus on actual problems and reducing the chances of manual error.

It is important to add at this point, that nearly 67% of the IT budget is operational and only 33% of the expense is of a capital nature. Since the operations involve a large expense, any advances in the process to improve availability or to bring down cost are of great value to an organization.

The network operations involve monitoring the vital parameters of the network to increase service availability uptime, improve service quality and mitigate outages, and monitor performance of the network devices for any potential signs that can cause an outage or security issues on the network. We will cover operations in more detail in Chapter 10, A Systematic Approach to Network Operations and Chapter 11, Basic Troubleshooting Skills and Techniques later in this book.

One needs to constantly evolve and improve to maintain a competitive advantage. A continual/continuous improvement process (CIP) is an ongoing effort to improve products, services, or processes. It is this desire of the organization to continuously evolve and improve that is addressed in the optimization phase.

This process is closely tied to the operations phase as the results of the operations phase are analyzed to detect recurring problems and to see if there are any design discrepancies, or enhancements that can be made to the network to improve service availability or performance. The goal of the optimize phase is to identify and resolve problems before they actually start to manifest themselves on the network resulting in service disruptions.

The optimize phase of the network lifecycle can cause changes to the network design to meet the service specifications with respect to functionality, performance, security, availability, and so on. In such cases, the network engineers go back to the drawing board to evaluate new alternatives and propose new approaches to meet the changing needs. Some of the things might need minor tweaks to the design, and hence can be handled by going back to the design phase of the network lifecycle, followed by implementation and testing. However, optimization can also be triggered by the maturity of a new technology in the market, and hence that needs a much broader scope, and the organization needs to start at the plan phase, and follow the complete cycle all over again. Whatever be the case, it is important to document the reasons for change, the proposed solution, and then follow it up with the rigor of the complete lifecycle in order for the network to be capable of meeting new requirements in a sustainable manner.