pune institute of computer technology department of

Pune Institute of Computer Technology

DEPARTMENT OF COMPUTER ENGINEERING

Vision

Achieve academic excellence through education incomputing, to create intellectual manpower to exploreprofessional, higher educational and social opportunities.

Mission

To impart learning by educating students with conceptual knowledge and hands on practices using modern tools, FOSS technologies and competency skills there by igniting the young minds for innovative thinking, professional expertise and research.

Computer Network Design and

Modeling

Course code: 410444B

By: Prof. Sudhakar

Reddy

8 August 2015 2PICT, Pune

Introduction, Requirement Analysis:

Concepts

Unit-I


Contents

• Overview of network analysis and design process

• System methodology

• System description

• Service description

• Service characteristics

• Performance characteristics

• Requirement analysis (user, application, device,

network, other) concepts

• Requirement specification and map


• What is CNDM?

• Why it is used?

• Why is it important?

• Where is it used?


Overview of network analysis and design

process

• Network analysis, architecture, and design are processes used

to produce designs that are logical, reproducible, and

defensible.

Network analysis

• Analysis involves learning what users, their applications and

devices need from the network.

• Purpose of analysis is for two reasons:

Listen to users and understand their needs

Understand the system


Fig: Inputs To and Outputs from the

network analysis process


Network architecture

• The network architecture process determines sets of technology and topology choices; the classes of equipment needed; and the relationships among network functions


Fig: Inputs To and Outputs

From the Network

Architecture Process

Network design

• Network design provides physical detail to the architecture.

• Physical detail includes blueprints and drawings of the

network; selections of vendors and service providers; and

selections of equipment (including equipment types and

configurations)


Fig: Inputs To and

Outputs From the

Network Design Process

Process Components


• Tactical and Strategic Significance

• Need to plan for the future and be ready for the changes.

• Current target (one year), Near-Term target (3 years) and Long-

Term target (5 years)


Fig: A one-/Three-/Five-year Project Plan


Fig: The Cyclic and Iterative Nature of Processes

•This cycle repeats for n number of times and will lead to long-

term target.

Hierarchy and Diversity


Fig: Hierarchy and Diversity in a network


Fig: Hierarchy added

Fig: Diversity added

Importance of Network analysis

• Understanding network and system complexity


Fig: Generations of networking


Fig: Hierarchy and

Traffic flow

Fig: Diversity added

to improve

performance of select

traffic flows

• Analysis helps us understand how technologies influence

networks, users, applications, and devices (and vice versa).

• Consider an example of adding a routing protocol to the

technology. Only analysis will let you know about the changes

and the implementations.

8 August 2015 PICT, Pune 17

Fig: Routing evolution

Architecture and Design defensibility

• Data from analysis, along with decisions made in architecture and design are documented. (Audit trail)

• These audit-trails can be provided to everyone who is using the network. Changes made can also be traced.

Model for Network Analysis, Architecture, and Design

• The model includes the following areas:

a. Defining the problems to be addressed

b. Establishing and managing customer expectations

c. Monitoring the existing network, system, and its environment

d. Analyzing data

e. Developing a set of options to solve problems

f. Evaluating and optimizing options based on various trade-offs

g. Selecting one or more options

h. Planning the implementation


A Systems Methodology

• Applying system’s methodology to networking means viewing

the network that you are architecting and designing, along with

a subset of its environment (everything that the network

interacts with or impacts), as a system.

• It includes interactions and dependencies between the network

and its users, applications, and devices.

• This methodology helps in determining , defining, and

describing the important characteristics and capabilities of the

network.


System Description

• How to define network as a system?

• System can be defined as a set of components that work

together to support or provide connectivity, communications,

and services to users of the system.


Fig: Generic components of a system

• Comparison of OSI model with the system


Fig: Comparison of OSI Layers to System Levels

• Subdividing the system into sub-components


Fig: Device component separated into constituents

• Reason for dividing the system into components is to

understand how these components interface with one

another.


Fig: A generic system with interfaces added

Service Description

• Description of the network services delivered to the system.


Fig: Grouping characteristics into

service levels and descriptions

Service Characteristics• These are individual network performance or functional

parameters that are used to describe services.

• Examples of service characteristics are:

Reliability

Manageability

Availability

Security

usability,

Flexibility

Capacity

Delay, etc.

• Measurements of these characteristics in the network to monitor, verify and manage services are called Service metrics.


Service Levels

• Service levels are the group of service characteristics to be

selected instead of selecting individual characteristics.

• Example: A service level (e.g., Premium) may combine

capacity (as 1.5 Mb/s) and reliability (as 99.99% )

• This is mostly a service provider view of the network.

• There are many ways to describe the service levels, such as

CIR: levels of capacity

CoSs: combine delay and capacity characteristics

ToSs: types of services

QoSs: quality of services

• Service levels can be applied to network as a system also.


Fig: Service requests, Offerings, and Metrics


System Components and Network Services

• System components like user, application and devices are

related to network services through requirements.

• Refined requirements of each component are called service

requirements and these are again refined to get network

services.


Fig: Requirements flow down components, from user

to network

Service Requests and Requirements

• Service requests are from the users, applications and devices.

• There are three types of service requests: Best effort

(unpredictable and unreliable), Predictable (predictable more

than best-effort) and guaranteed services (predictable and

reliable).

• For this to be achieved, better understanding of requirements is

necessary.

• E.g: A device requires capacity (bandwidth) between 4 and 10

Mb/s. You have to determine a method of controlling the

information flow and network resources to maintain this

bandwidth.

• Capacity is a finite resource in the network.


• Example of capacity in terms of accepting calls ( can be solved

in best-effort or guaranteed service).

• Best-effort, predictable and guaranteed services refer to the

degree of predictability of a request.

• Networks can be low or high performance networks based on

their delivery of the above services.

• Conclusion: Network services are requested by the users,

applications and devices in terms of requirements.

• These services are low or high performance, based on their

attainment of requests.


Service offerings

• The service given by the network to the users, applications and

devices.

• Services which are unpredictable get the best-effort service

offering.

• Services which are predictable or limited get the guaranteed or

predictable service offerings.

• Note: Guaranteed services are not always of high

performance and best-effort services are not always of low

performance.


Service metrics

• Services which cannot be configurable, measurable and

verifiable are of no use and to do it, we have metrics.

• Service metrics will help us in defining threshold and limit.


Fig: Example of

threshold and limit

Performance Characteristics

• Performance characteristics we study here are capacity, delay

and RMA.

• Capacity is label to define class of characteristics like

bandwidth, throughput, goodput, etc.

• Delay is label for end-to-end delay, round trip delay, and delay

variation.

• RMA is a label for reliability, maintainability, and availability.

Capacity

• It is a measure of system’s ability to transfer information.

• Examples of Capacity: bandwidth, throughput, goodput, etc.


Delay

• Delay is a measure of the time difference in the transmission

of information across the system.

• There can be various sources of delay like propagation,

transmission, queuing and processing.

• Measured in one direction (end-to-end) and both directions

(round-trip).

• Processing delay is called latency for devices and applications.

• Delay variation is the change in delay over time. Also known

as Jitter.


RMA

• Reliability

• It is the statistical indicator of the frequency of failure of the

network and its components and represents the unscheduled

outages of the service. It gives some degree of predictable

behavior.

• Maintainability

• It is a statistical measure of the time to restore the system to

fully operational status after it has experienced a fault.

Generally expressed as a mean-time-to-repair (MTTR).

• Availability

• Availability is the relationship between the frequency of

mission-critical failures and the time to restore service.

• A= (MTBCF)/(MTBCF + MTTR) or A = (MTBF)/(MTBF + MTTR)


Performance Envelopes

• A performance envelope is a combination of two or more

performance requirements, with thresholds and upper and/or

lower limits for each.

• Within this envelope, levels of application, device, and/or

network performance requirements are plotted.


Fig: 2D and 3D performance envelopes

Network supportability

• It is a mistake to assume that a successful network architecture

and design meet the requirements only on the day it is

delivered to the customer and that future requirements are the

responsibility of the customer.

• Operations and support constitutes 80% of the life-cycle costs

of a system, whereas development, acquisition, and

installation represent only 20%.

• Post implementation phases can be broken down into 3

elements: Operations, maintenance and human knowledge.

• Operations- identifying actions

• Maintenance-accomplishing the actions

• Human knowledge- documenting and training the personnel.


• Key characteristics of a network architecture and design that affect the post implementation costs include:

– Network and system reliability

– Network and system maintainability

– Training of the operators to stay within operational constraints

– Quality of the staff required to perform maintenance actions

• Some examples of key network architecture/design decisions that affect these characteristics include:

– Degree of diversity in component selection, Quality of network components, Location and accessibility of components

• Two major tasks must be accomplished to ensure supportability:

– Validated Conformance and documentation of non-conformance (at-least)

– Maintenance personnel must be properly trained.


Requirement analysis: concepts

• Requirements obtained from many sources are categorized into

sub-requirements like functional, feature, rejected, etc.


Fig: Requirements Are Separated into Core/Fundamental Requirements,

Features, Future, Rejected, and Informational Requirements

Categorizing the requirements

• Based on the practice of IETF-RFC 2119. we have identifying

key words and phrases.

1. Must/Shall/Required- fundamental requirements/ core

2. Must Not/Shall Not- restriction/prohibition (core)

3. Should/Recommended- not absolutely necessary

(feature/future requirement)

4. Should Not/Not Recommended- not absolutely necessary and

restriction (feature/future requirement)

5. May/Optional- truly optional, can be a feature/future or

rejected.


User requirements

• User refers not only to the end users but everyone involved in the system.

• Lower the level, more technical will be the requirements.


Fig: Types of User Requirements

1. Timeliness- tolerable time frame by the user. Measurement:

End-to-end delay or round-trip delay

2. Interactivity- response time of the system and network. E.g.-

file download. Measurement: Round-trip delay

3. Reliability- availability in user’s perspective. He should be

able to have access to resources most of the time and

consistent service. Measurement: reliability, delay and

capacity

4. Presentation Quality- quality of the presentation to the user.

E.g.- audio, video and displays. Measurement: all

performance characteristics

5. Adaptability- ability of the system to adapt to users’

changing needs. E.g.- distance independence and mobility.

Measurement: ----


6. Security- a requirement to guarantee the confidentiality,

integrity, and authenticity of a user’s information and physical

resources. Measurement: reliability, capacity and delay

7. Affordability- purchases fit within a budget. Measurement:

budget (non-technical)

8. Functionality- any functional requirement that user has for

the system.

9. Supportability- a set of characteristics that describes how

well the customer can keep the network operating at designed

performance. How the staff and NOC will support the

changes in the network?

10. Future growth- If the users are planning to deploy and use

new applications and devices in the network.


Application Requirements

• Requirements from application information, experience, or

testing, and represent what is needed by the applications to

successfully operate on the system.


Fig: Types of Application Requirements

Types of Applications

1. Mission-critical applications- have predictable, guaranteed,

and/or high performance RMA requirements

Examples: transactions and money handling, airline

reservation, credit card processing, telemetry, billing apps,

health-care monitoring, etc.

2. Rate-critical applications- have predictable, guaranteed,

and/or high-performance capacity requirements

Examples: voice, non-buffered video, telemedicine,

teleseminars, teleconferencing, etc.

3. Real-time and interactive applications- have predictable,

guaranteed, and/or high performance delay requirements

Examples: non-buffered video, ATM transactions-> real time

Telnet, FTP, web apps-> interactive


Delay types

• Interactive burst applications are those where network delays are predominant than application/device delays. Ex: telnet

• Interactive bulk applications are those where application/device delays are predominant than network delays. Ex: FTP (large file)


Fig: Delay types

Application Groups

• It is easy to group the applications because every time you have the similar requirements, choosing the group of related application and other requirements becomes easy.

1. Telemetry/Command-and-Control Applications- ATM, controlling spacecraft, sensors at home, etc.

Real time and/or interactive delay and mission critical

2. Visualization Applications- aeronautics, weather modeling, molecular simulations, gaming, etc.

Rate critical and interactive burst

3. Distributed-Computing Applications- cloud services, apps in LAN, etc.

Real-time or interactive burst

4. Web Development, Access, and Use Applications- Telnet, FTP, etc.

Interactive (mix of interactive burst and bulk)


5. Bulk Data Transport Applications- FTP for large files,

MFTP and ARCP, etc.

Do not have high performance requirements

6. Tele∗Service Applications- telemedicine, teleseminars,

teleconferencing, etc.

Interactive & Real-time delay and rate critical &

mission-critical

7. Operations, Administration, Maintenance, and

Provisioning (OAM&P) Applications- DNS, SMTP, address

resolution service, network monitoring, etc.

Mission critical and interactive

8. Client–Server Applications- ERP, supply chain management

(SCM), CRM, etc.

Mission critical and interactive


Application Locations

• Location of the application can also be determined.

• Few applications are on all the systems

• Few apps are on different systems.

• Need to draw a map showing all the apps at different locations.


Fig: An example of Applications Map

• Requirements from devices that network will support,

particularly the types of devices, their performance

characteristics, and their location information.


Device Requirements

Fig: Types of Device Requirements

Device types

• Grouped into three categories

1. Generic computing devices

2. Servers

3. Specialized devices

1. Generic computing devices

• Includes all desktop. Laptop and handheld devices which

typically serve a single user.

• Template for types of devices under generic devices can be

seen in the next slide.

• These devices requirements are important


• Note: Last Foot Problem- getting services and performance from the devices’ network interface through the device to its applications and users.


Fig: An Example template for device description

2. Servers

• Servers are computing devices that provide a service to one or

more users (clients).

• More processing, more power, more storage, more

peripherals, more networking, etc.

• Need more attention on requirements than generic devices.

3. Specialized devices

• Specialized devices are devices that provide specific

functions to their users.

• Generally do not support direct access to user applications,

but gather, produce, or process information to be sent to

users.

• Examples: super computers, parallel or distributed computing

systems, medical devices, networked cameras or tools, etc.


• If a network is to be designed for an ATM machine, library, medical center, universities, research centers, then the devices will be location dependent as well as the network.


Fig: Specialized devices

Performance characteristics

• It is better to understand the performance characteristics of the devices by testing them.

• Sometimes, issues of devices are misinterpreted as network issues.

• All the components of a devices should be working effectively to support the network and if not, network shou8ld consider them.


Fig: Device components

• Following performance characteristics to be considered:

1. Storage performance

2. Processor (CPU) performance

3. Memory performance (access times)

4. Bus performance (bus capacity and arbitration efficiency)

5. OS performance

6. Device driver performance

• On analyzing the requirements of these characteristics,

bottlenecks can be identified in the early stages.


Device locations

• Location information helps to determine the relationships

among components of the system.


Fig: Device locations

Network Requirements

• Requirements from the existing networks and constraints

should be considered for a proper network design to avoid

glitches in future.


Fig: Types of Network requirements

Existing Networks and Migration

1. Scaling dependencies- adding new network to the old

2. Location dependencies- locations constraints of existing

network

3. Performance constraints- issues with existing network

performance characteristics

4. Network, system, and support service dependencies- various

dependencies

5. Interoperability dependencies- design dependencies of

existing networks

6. Network obsolescence- devices which are at the verge of their

usefulness.


Network Management and Security

• 4 categories of network management tasks:

1. Monitoring for event notification

2. Monitoring for metrics and planning

3. Network configuration

4. Troubleshooting

• Existing networks are monitored for a period of time.

• Where parameters from network devices, data, operations and storage are obtained and issues are troubleshoot-ed.

• Few of the n/w management requirements are:– Monitoring methods

– Instrumentation methods. These include the network management protocols (SNMPv3, CMIP, RMON), parameter lists (MIBs), monitoring tools, and access methods

– Sets of characteristics for monitoring

– In-band versus out-of-band monitoring

– Centralized versus distributed monitoring

– Performance requirements



Fig: Security Risk Assessment

Other Requirements

Supplemental Performance Requirements

• Characteristics after the implementation of the network. They are: Operational suitability (how well it is configured, monitored and

adjusted by customers), supportability (how well the customer can keep

the performance) and confidence (ability of the n/w to deliver).

Financial Requirements

• The financial requirements gathered during the analysis process will be combined with the users’ affordability requirements, to form a complete financial picture of the network

Enterprise Requirements

• Integration of phone, FAX, voice and video types of requirements over the same transmission infrastructure as data is becoming common.


The Requirements Specification and Map

• Requirements specification is a document that lists and

prioritizes the requirements gathered for your architecture and

design

• Requirements map shows the location dependencies between

applications and devices.

• Requirements need to be gathered from variety of sources,

existing network, applications, devices, etc.

• Verbal, derived, and have to be estimated.

• Determine whether they are core/functional, feature, rejected,

etc.

• Priority levels of funds to be spent, etc.


• Fields of the requirement specification template

1. ID/Name- identifier of the requirement or the name.

2. Date- the day it was developed.

3. Type- component of the requirement (User, App, Device, N/w, Other).

4. Description- details of the requirement. Indicate functional or performance.

5. Gathered/Derived- details of origin of the requirement.

6. Locations- where the req. applies in the environment.

7. Status- current state of the req. (core, feature, future, rejected, etc)

8. Priority- number of the priority level.


Fig: Template for the Requirements Specification

• Example of requirements

A first attempt was made to gather requirements for building a LAN network.

The results were as follows:

1. 150 users (60 engineers, 15HR and Finance, 30 Manufacturing, 10 Management, 30 Sales/Marketing, 5 Other).

2. Each area in the building must support Fast Ethernet connections to the backbone.

3. Database, Visualization, Manufacturing, and Payroll applications are considered mission-critical for this company.

4. Inventory application (INV1) for manufacturing requirements not determined at this time.

5. Database application (DB1) requires a minimum of 150 Kb/s, per session.

6. Engineering users have workstations with GigE NICs.

7. Visualization application (VIS1) for finance requires up to 40 Mb/s capacity and 100 ms round-trip delay.


8. Payroll application (PAY1) requires 100% uptime (while in

operation) between finance and outside payroll company.

9. Company must be kept secure from Internet attacks.

10. Company requires a minimum of T1 access to Internet.

11. Current network will be completely replaced, so there are no

requirements from existing network.

12. Other general applications: mail, word processing, internal

and external Web access.



Fig: The

Beginning of

a Requirement

Specification


Fig: The Beginning of a Requirements Map

END of Unit-I


pune institute of computer technology department of

Documents