Modern networks and applications demand improvements in network performance management practices. The Network Life Cycle (NLC) provides a comprehensive framework against which network managers can make critical decisions, whether in moments of immediate crisis or for long-range planning.
The NLC defines a range of needs that require a variety of methodologies and tools for network performance management through Monitoring, Measurement, Assessment and Diagnosis (MMAD). Each phase of the NLC has its own requirements that can typically be met with a judicious selection of tools and assessment solutions.
Figure 1 – the Network Life Cycle
The major phases are:
- Business Case
- Request for Proposals
Each phase of the typical Network Life Cycle requires some degree of attention to MMAD. In fact, the degree to which each phase is properly carried out defines the likelihood of success of subsequent phases.
To illustrate the importance of each phase of the Cycle, consider that some networks are constructed with little or no Staging phase -- the Deployment phase from one service to another is made "over the weekend" with fingers crossed. The success of the deployment is dependent on a particularly large expenditure in the Planning phase -- and the risk for lack of a thorough Staging effort is assumed by Deployment and even Operation. This may be reasonable since the cost of a Staging phase can be considerable, both in terms of money and time expended.
Let's look more closely at the details of each phase of the NLC to identify the tasks involved:
- Identify needs
- Define advantage
- Define objectives
- Establish risk tolerance
- Assign budget
- Define scope and bounds
- Create base of use-cases
- Establish technical criteria
- Detail high-level requirements
- Generate performance specifications
Request for Proposals
- Review technologies
- Design solution(s)
- Do risk analysis
- Generate proof-of-concept
- Select implementation target
- Define project
- Assign responsibilities/roles
- Define milestones/deliverables
- Assign resources
- Identify dependencies
- Deployment/cutover strategies
- Contingency planning
- Implement test network
- Perform simulation
- Perform stress tests
- Apply range-of-use applications
- Identify installation/configuration issues
- Identify maintenance issues
- Pre-qualification in situ
- Stress testing final installation
- Reality check (next morning)
- Recovery/fallback (as required)
- Maintenance and upgrading
- On-going support
- Compare performance to plan
- Review timeline of needs
- Projected needs/provisioning
- Proposed revisions against changed requirements
- Growth strategy
The specific MMAD tools and methodologies for each stage tend to overlap but are typically distinct. The balance of this guide offers a range of answers that deal with particular tasks within the NLC framework.
As you consider the recommended best practices, tools, and methodologies, give due consideration to more than just your day-to-day requirements. Think broadly about the nature of today's networks and how each solution applies in your case. Aligning your network performance management practices with overall NLC objectives can make your life easier, expenditures more accountable, and your success more likely.
With a Ph.D. in computational physics from McGill University, Loki Jorgenson has been active in scientific computation, physics and mathematics, visualization, and simulation for over 20 years. He has published in areas as diverse as philosophy, graphics, educational technologies, statistical mechanics, logic and number theory. Dr. Jorgenson is an Adjunct Professor of Mathematics at Simon Fraser University where he co-founded the Center for Experimental and Constructive Mathematics. He has directed academic research in numerous network-centric projects from high performance computing to telelearning, working closely with private sector partners and government. At Apparent Networks, Dr. Jorgenson is responsible for leading their efforts in critical research areas such as high performance applications, wireless and VoIP, expert systems, and intelligent networks in collaboration with international thought leaders and as a corporate member of Internet2.
This was first published in June 2004