Ch14

1

Chapter 14

Network Design and Management

Data Communications andComputer Networks: A Business User’s Approach

2

Data Communications and Computer Networks Chapter 14

IntroductionProperly designing a computer network is a difficult task. It requires planning and analysis, feasibility studies, capacity planning, and baseline creation skills.

Performing network management is difficult too. A network manager must possess computer and people skills, management skills, financial skills, and be able to keep up with changing technology.

3


Systems Development Life Cycle (SDLC)Every business has a number of goals.

System planners and management personnel within a company try to generate a set of questions, or problems, to help the company achieve those goals.

To properly understand a problem, analyze all possible solutions, select the best solution, and implement and maintain the solution, you need to follow a well-defined plan.

SDLC is a methodology, or plan, for a structured approach to the development of a business system.

4

Systems Development Life CycleSDLC involves several phases. These phases are often:

• Planning

• Analysis

• Design

• Implementation

• Maintenance

These phases are cyclical and usually never ending.

5


Cycle of the phases

6


Systems Development Life CycleA systems analyst is typically responsible for managing a project and following the SDLC phases.

Anyone, however, may be called upon to assist a systems analyst.

Or anyone may have to assume some of the duties of a systems analyst.

Individuals that are called upon to support a computer network should understand the basic phases of SDLC.

7


Systems Development Life CyclePlanning Phase - Identify problems, opportunities, and objectives.

Analysis Phase - Determine information requirements. Information requirements can be gathered by sampling and collecting hard data, interviewing, questionnaires, observing environments, and prototyping.

Design Phase - Design the system that was recommended and approved at the end of the analysis phase.

8


Systems Development Life CycleImplementation Phase - The system is installed and preparations are made to move from the old system to the new.

Maintenance Phase - The longest phase, involves the ongoing maintenance of the project.

Maintenance may require personnel to return to an earlier phase to perform an update.

9


10


Network ModelingWhen updating or creating a new computer system, the analyst will create a set of models for both the existing system (if there is one) and the proposed system.

Network models can either demonstrate the current state of the network or can model the desired computer network.

A location connectivity diagram is a network modeling tool that depicts the various locations involved in a a network and the interconnections between those locations.

11


Network ModelingAn overview location connectivity diagram shows the big picture of geographic locations of network facilities.

External users and mobile users can be identified, as well as the locations primary to a business.

A detailed location connectivity diagram is a close-up model of a single location and the networks that reside at the location.

Working groups and the distances between those groups can be identified with a detailed diagram.

12


X means a special site

13


14

Feasibility StudiesThere are a number of ways to determine if a proposed system is going to be feasible.

Technically feasible means the proposed system can be created and implemented using currently existing technology.

It can be done.

Financially feasible means the proposed system can be built given the company’s current financial ability.

We can afford it.

Operationally feasible means the system operates as designed and implemented.

It’ll work after we build it and it will be used.

15

Feasibility StudiesTime feasible means the system can be constructed in an agreed upon time frame.

Payback analysis ascertains costs and benefits of proposed system usually on an annual basis.

Payback analysis is a good technique to use to determine financial feasibility.

To calculate payback analysis, you must know all the expenses that will be incurred to create and maintain the system, as well as all possible income derived from the system.

You must also be aware of the time value of money (a dollar today is worth more than one dollar promised a year from now because the dollar can be invested).

16

System Costs:One time costs:

Personal costs

Computer usage costs for analysis

Hardware and software

Training, support and management

Supplies, furniture, space

Recurring costs:

Lease payments

Licenses

Salaries of support personal

Ongoing supplies

Maintenance

Planned replacement

17


18

Note profitat 6th year

19


Capacity PlanningCapacity planning involves trying to determine the amount of network bandwidth necessary to support an application or a set of applications.

A number of techniques exist for performing capacity planning, including linear projection, computer simulation, benchmarking, and analytical modeling.

Linear projection involves predicting one or more network capacities based on the current network parameters and multiplying by some constant.

20


Capacity PlanningA computer simulation involves modeling an existing system or proposed system using a computer-based simulation tool.

Benchmarking involves generating system statistics under a controlled environment and then comparing those statistics against known measurements.

Analytical modeling involves the creation of mathematical equations to calculate various network values.

21


Creating a BaselineInvolves the measurement and recording of a network’s state of operation over a given period of time.

A baseline can be used to determine current network performance and to help determine future network needs.

Baseline studies should be ongoing projects, and not something started and stopped every so many years.

22


Creating a BaselineTo perform a baseline study, you should:

• Collect information on number and type of system nodes, including workstations, routers, bridges, switches, hubs, and servers.

• Create an up-to-date roadmap of all nodes along with model numbers, serial numbers and any address information such as IP or Ethernet addresses.

• Collect information on operational protocols used throughout the system.

23


Creating a BaselineTo perform a baseline study, you should:

• List all network applications, including the number, type and utilization level.

• Create a fairly extensive list of statistics to help meet your goals. These statistics can include average network utilization, peak network utilization, average frame size, peak frame size, average frames per second, peak frames per second, total network collisions, network collisions per second, total runts, total jabbers, total CRC errors, and nodes with highest percentage of utilization.

24


25


Network Manager SkillsA good network manager will possess many skills:

• Computer skills

• People skills

• Management skills

• Financial planning skills

• Knowledge of statistics

• Speaking and writing skills

26

Generating Useable StatisticsStatistics, properly generated, can be an invaluable aid to demonstrating current system demands and predicting future needs.

Mean time between failures (MTBF)

Mean time to repair (MTTR)

Availability is the probability that a particular component or system will be available during a fixed time period

Reliability is the probability that over a period of time the particular component or device will be available

27

Generating Useable StatisticsMean time between failures (MTBF) is the average time a device or system will operate before it fails.

(from the manufacturer!)

Large or small?

Mean time to repair (MTTR) is the average time necessary to repair a failure within the computer system.

Includes swap time, time to bring the system back up, replacement time, etc.

Large or small?

28

Generating Useable StatisticsAvailability is the probability that a particular component or system will be available during a fixed time period.

A(t) = a/(a+b) + b/(a+b) x e-(a+b)t A = 1 means the system is nearly always operational

in which: a = 1/MTTRb = 1/MTBFe = natural log functiont = the time interval

29

Generating Useable StatisticsSuppose we want to calculate the availability of a modem that has a MTBF of 3000 hours and a MTTR of 1 hour. The availability of this modem for an 8-hour period is:

a = 1/1b = 1/3000 = 0.00033A(8 hours) =1/(1 + 0.00033) + 0.00033/(1 + 0.00033) x e-(1 + 0.00033)8

= 0.9997 + 0.00033 x 0.000335= 0.9997

Not available 3 out of 10,000 times you want it.

30


Generating Useable Statistics

Reliability is the probability that over a period of time the particular component or device will be available (not fail):

R(t) = e –bt

R(0) = e –0 = 1

in which: b = 1/MTBFt = the time interval of the operation

31



What is the reliability of a modem if the MTBF is 3000 hours and a transaction takes 20 minutes, or 1/3 of an hour (0.333 hours):

R(0.333 hours) = e -(1/3000)(0.333) = e -0.000111 = 0.99989

Not reliable for .011 percent of the time.

32



So what do you want?

Availability and reliability between 0.9999 and 0.99999 is desired!

What is this in number of hours between failure for a year of service?

33


Managing OperationsThere are many services and functions available to assist an individual in managing computer network operations.

One of the more useful is Simple Network Management Protocol (SNMP).

SNMP is an industry standard designed to manage network components from a remote location.

Currently in version 3, SNMP supports agents, managers, and the Management Information Base (MIB).

34


Managing OperationsA managed element has management software, called an agent, running in it.

A second object, the SNMP manager, controls the operations of a managed element and maintains a database of information about all managed elements.

A manager can query an agent to return current operating values, or can instruct an agent to perform a particular action.

The Management Information Base (MIB) is a collection of information that is organized hierarchically and describes the operating parameters of all managed agents.

35


Managing OperationsSNMP operates on a network between the application layer and the UDP/IP layer (not TCP/IP) – the transport layer.

36


Network Diagnostic ToolsTo assist a network support person, a number of diagnostic tools are available:

• Electrical testers (voltage, etc.)

• Cable testers (open circuits, etc.)

• Network testers (view the entire network0

• Protocol analyzers (monitors packets)

37


Capacity Planning and Network Design In Action: BringBring CorporationReturning to BringBring Corporation from an earlier chapter, let’s complete our design, including e-mail and Internet access for each of the four sites.

A linear projection can be used to estimate the amount of Internet traffic at each site.

An overview location connectivity diagram gives us a big picture of the network interconnections.

38


39


Capacity Planning and Network Design In Action: BringBring CorporationA second linear projection can be used to determine the amount of local area network traffic within each site.

Technology

Ch14