Transcript Network Modelling

Network Modeling
Introduction

The chapter will address the following questions:





Why may network modeling become an important skill for
applications developers in the next several years?
What is the description network modeling and explain why it is
important?
What is the definition of a system in terms of locations, location
types, and clusters?
How can you factor a system’s or application’s locations into
component locations using a special location decomposition
diagram?
How can you document the connections and essential data flows
between locations using location connectivity diagrams (LCDs)?
1
Network Modeling
Introduction

The chapter will address the following questions:



What is the complementary relationship between network, process,
and data models?
How can you synchronize data, process, interface, and network
models to provide a complete and consistent logical system
specification?
How is network modeling useful in different types of projects and
phases?
2
Network Modeling
Network Modeling - Not Just For Computer
Networks

Computer Networks


Have become the nervous system of today’s information systems.
The computer network is a physical component of an information
system.
 Must be created to support the logical distribution of data,
processes, and interfaces of an information system.
 Network modeling is a technique for documenting the
geographic structure of a system. Synonyms include
distribution modeling and geographic modeling.
3
Network Modeling
INFORMATION SYSTEMS FRAMEWORK
FOCUS ON
SYSTEM
DATA
FOCUS ON
SYSTEM
PROCESSES
FOCUS ON
SYSTEM
GEOGRAPHY
FAST
Methodology
Operating Locations
Survey Phase
(establish scope &
project plan)
FOCUS ON
SYSTEM
INTERFACES
SYSTEM
OWNERS
(scope)
Location Decomposition
Data Requirements
Business Processes
Interface Requirements
Communication Reqts.
E DI
Cust
S
Y
S
T
E
M
A
N
A
L
Y
S
T
S
or der
S t.
Louis
HQ
ship
or der
W est
Custom er s
SYSTEM
USERS
catalog
changes
P r oducts
Catalog
E ast
Custom er s
cr edit
cr edit
LA
Office
(requirements)
Study Phase
(etablish system
improvemetn
objectives)
ship
or der
Indy
W ar ehouse
ship or der
NY
Office
ser vice
Maintenance
Recor ds
Logical Data Model
Logical Process Model
Logical Interface Model
Location Connectivity
Diagram
SYSTEM
DESIGNERS
(specification)
SYSTEM
BUILDERS
(components)
Database
Technology
(and standards)
Software
(and Hardware)
Technology
Interface
Technology
Networking
Telchnology
(and standards)
(and standards)
(and standards)
4
Definition Phase
(establish and
prioritize
business system
requirements)
Network Modeling
Network Modeling - Not Just For Computer
Networks

Computer Networks

The need for network modeling is being driven by a technical trend
– distributed computing.
 Distributed computing is the assignment of specific
information system elements to different computers which
cooperate and interoperate across computer network. A
synonym is client/server computing; however, client/server is
actually one style of distributed computing.
 The distributed computers include:
• desktop and laptop computers, sometimes called clients
• shared network computers, called servers
• legacy mainframe computers and minicomputers
5
Network Modeling
System Concepts For Network Modeling

Today’s systems analyst must seek answers to new
questions:






What locations are applicable to this information system or
application?
How many users are at each location?
Do any users travel while using (or potentially using) the system?
Are any of our suppliers, customers, contractors, or other external
agents to be considered locations for using the system?
What are the user’s data and processing requirements at each
location?
How much of a location’s data must be available to other
locations? What data is unique to a location?
6
Network Modeling
System Concepts For Network Modeling

Today’s systems analyst must seek answers to new
questions: (continued)



How might data and processes be distributed between locations?
How might data and processes be distributed within a location?
A network modeling tool is needed to document what
we learn about a business system’s geography and
requirements.

Network modeling is a diagrammatic technique used to document
the shape of a business or information system in terms of its
business locations.
7
Network Modeling
System Concepts For Network Modeling

Business Geography



Logical network modeling is the modeling of business network
requirements independent of their implementation.
All information systems have geography.
The location connectivity diagram (LCD) models system
geography independent of any possible implementation.
 A location connectivity diagram (LCD) is a logical network
modeling tool that depicts the shape of a system in terms of its
user, process, data, and interface locations and the necessary
interconnections between those locations.
8
Network Modeling
Buyers
(20)
on-the-road
Suppliers
(275)
Purchasing
Clerks
(3)
Atlanta
Accounts
Payable
Office
(Atlanta)
Inventory Control
Manager's Office
(Atlanta)
Central
Warehouse
(Atlanta)
Distribution
Center
(New York)
Distribution
Center
(Chicago)
9
Distribution
Center
(Los Angeles)
Network Modeling
System Concepts For Network Modeling

Business Geography

The location connectivity diagram (LCD) illustrates two concepts –
locations and connectivity.
 The concept of geography is based on locations.
• A location is any place at which users exist to use or interact with
the information system or application. It is also any place where
business can be transacted or work performed.
Location
Business management and users will tend to identify logical
locations where people do work or business.
 Information technologists will tend to discuss physical
locations where computer and networking technology is
located.

10
Network Modeling
System Concepts For Network Modeling

Business Geography

Example locations include:
Logical locations – places where data are
collected, work is performed, or
information is needed
 City
 Campus
 Building
 Office
 Work area (e.g., warehouse)
 Subsidiary
 Home office
 Customer, supplier, or contractor
11
Implementation Locations – places
where computers, peripherals, and other
information technology is located
 Computer center
 Network server
 PC or terminal location
 Local area network
 Wide area network hub/gateway
Network Modeling
System Concepts For Network Modeling

Business Geography


Logical locations can be:
 scattered throughout the business for any given information
system.
 on the move (e.g., traveling sales representatives).
 external to the enterprise for which the system is being built.
For instance, customers can become users of an information
system via the telephone or the Internet.
Logical locations can represent:
 clusters of similar locations
 organizations and agents outside of the company but which
interact with or use the information system; possibly (and
increasingly) as direct users
12
Network Modeling
System Concepts For Network Modeling

Business Geography

Specific
Location
Cluster
of "like"
locations
Moving or Mobile
Location(s)
External
Location
Derivatives of the rectangle will be used to illustrate different types
of locations.
 The standard rectangle will be used to represent a specific
location.
 The rectangle with the double, vertical lines will be used to
represent a cluster of locations.
 Some locations are not stationary, a rounded rectangle will
represent their mobility.
 Some locations represent external organizations and agents
(such as customers, suppliers, taxpayers, contractors, and the
like). A parallelogram to illustrate these external locations.
13
Network Modeling
System Concepts For Network Modeling

Business Geography


Location names should describe the location and/or its users.
Examples of location names follows:
 Paris, France
Indianapolis, Indiana
Grissom Hall
 Building 105
Grant Street building
Room 222
 Warehouse
Rooms 230-250
Shipping Dock
 Order Clerk
User names (as locations) Order Entry Dept.
 Customers
Order clerks (a cluster)
Suppliers
 Students
14
Network Modeling
System Concepts For Network Modeling

Business Geography



Some locations consist of other locations and clusters.
It can be quite helpful to understand the relative decomposition of
locations and types of location.
 Decomposition is the act of breaking a system into its
component subsystems. Each ‘level’ of abstraction reveals
more or less detail (as desired) about the overall system or a
subset of that system.
In systems analysis, decomposition allows you to partition a
system into logical subsets of locations for improved
communication, analysis, and design.
 A location decomposition diagram shows the top down
geographic decomposition of the business locations to be
included in a system.
15
Network Modeling
Inventory
Control
System
Geography
Atlanta
Headquarters
Inventory
Control
Manager
Purchasing
Agents
(4)
Accounts
Payable
Managers
(2)
New York
Distribution
Center
Accounts
Payable
Office
Los Angeles
Distribution
Center
Buyers
(15-25)
Accounts
Payable
Clerks
(3)
16
Chicago
Distribution
Center
Network Modeling
System Concepts For Network Modeling

Business Geography




The purpose of network modeling is to help system designers
distribute the technical data, processes, and interfaces across the
computer network.
The systems analyst needs to specify the technology-independent
communications that must occur between business locations.
The communication between business locations requires
connectivity.
 Connectivity defines the need for, and provides the means for
transporting essential data, voice, and images from one location
to another.
Connections between locations represent the possibility of data
flows between locations.
17
Network Modeling
Buyers
(15-25)
Chicago
Distribution
Center
Purchasing
Agents
(4)
Los Angeles
Distribution
Center
New York
Distribution
Center
Suppliers
Accounts
Payable
Managers
(2)
Accounts
Payable
Clerks
(3)
18
Inventory
Control
Manager
Network Modeling
System Concepts For Network Modeling

Miscellaneous Constructs

There are no universal standards for location connectivity
diagrams; therefore, in appropriate situations it is permissible to
annotate LCDs with symbols from other models, such as data flow
diagrams.
19
Network Modeling
System Concepts For Network Modeling

Synchronizing of System Models


Network, data, interface, and process models represent different
views of the same system, but these views are interrelated.
Modelers need to synchronize the different views to ensure
consistency and completeness of the total system specification.
20
Network Modeling
System Concepts For Network Modeling

Synchronizing of System Models

Data and Process Model Synchronization:
 There should be one data store in the process models for each
entity in the data model. Also, there are sufficient processes in
the process model to maintain the data in the data model.
 The synchronization quality check is stated as follows:
• Every entity should have at least one C, one R, one U, and one D
entry for system completeness. If not, one or more event processes
were probably omitted from the process models. More importantly,
users and management should validate that all possible creates,
reads, updates, and deletes have been included.
21
R
R
R
.Customer Name
C
C
U
R
R
R
.Customer Address
C
C
U
RU
RU
RU
R
R
.Customer Credit Rating
C
U
.Customer Balance Due
Order
R
RU
U
R
R
C
D
RU
RU
.Order Number
C
R
R
.Order Date
C
U
U
.Order Amount
C
U
U
CRUD
CRUD
CRUD
CRUD
C
Ordered Product
.Quantity Ordered
C
.Ordered Item Unit Price
C
D
Process Product Inventroy
Adjustment
Process Internal Change to
Customer Order
R
R
Process Change to Product
Specification
Process Customer Change to
Outstanding Order
R
R
Process Product Price Change
Process Customer Order
Cancellation
R
C
Process Product Withdrawl from
Market
Process New Customer Order
C
C
Process New Product Addition
Process Customer Credit
Application
C
.Customer Number
Entity . Attribute
Customer
Process Customer Change of
Address
Process Customer Application
Process Internal Customer Credit
Change
Network Modeling
RU
RU
RU
RU
CRUD
CRUD
R
R
R
R
C
.Product Number
R
R
R
R
C
R
.Product Name
R
R
R
C
RU
.Product Description
R
R
R
C
.Product Unit of Measure
R
R
R
C
.Product Current Unit Price
R
R
R
.Product Quantity on Hand
RU
RU
RU
Product
22
U
D
RU
RU
RU
U
RU
Network Modeling
System Concepts For Network Modeling

Synchronizing of System Models

Data and Network Model Synchronization:
 A data model describes the stored data requirements for a
system as a whole.
 The network model describes the business operating locations.
 The goal is to identify what data is at which locations.
 Specifically, the following business questions might be asked:
• Which subset of the entities and attributes are needed to perform
the work to be performed at each location?
• What level of access is required?
• Can the location create, read, delete, or update instances of the
entity?
23
Network Modeling
System Concepts For Network Modeling

Synchronizing of System Models

Data and Network Model Synchronization: (continued)
 System analysts have found it useful to define logical
requirements in the form of a Data-to-Location-CRUD matrix.
• A Data-to-Location-CRUD Matrix is a table in which the rows
indicate entities (and possibly attributes); the columns indicate
locations; and the cells (the intersection rows and columns)
document level of access where C = create, R = read or use, U =
update or modify, and D = delete or deactivate.
24
.Customer Name
R
CRUD
R
CRUD
R
CRUD
R
.Customer Address
RU
R
CRUD
R
CRUD
R
CRUD
R
.Customer Credit Rating
X
R
RU
R
.Customer Balance Due
R
R
RU
R
INDV
ALL
.Order Number
SRD
R
.Order Date
SRD
.Order Amount
. Warehose
R
RU
.Customer Number
San Diego
CRUD
. Sales
. Warehouse
SS
R
San Francisco
. Sales
SS
CRUD
INDV
Boston
. Sales
SS
R
. Advertsing
SS
CRUD
. Marketing
ALL
R
Kansas City
ALL
R
Entity . Attribute
Customer
Customers
. Warehouse
. Accounts
……..Receivable
Network Modeling
R
R
SS
ALL
CRUD
R
CRUD
R
CRUD
R
SRD
R
CRUD
INDV
ALL
.Quantity Ordered
SUD
R
CRUD
.Ordered Item Unit Price
SUD
R
CRUD
ALL
ALL
ALL
ALL
ALL
ALL
ALL
ALL
ALL
.Product Number
R
CRUD
R
R
R
R
R
R
R
.Product Name
R
CRUD
R
R
R
R
R
R
R
.Product Description
R
CRUD
RU
R
R
R
R
R
R
.Product Unit of Measure
R
CRUD
R
R
R
R
R
R
R
.Product Current Unit Price
R
CRUD
R
R
R
R
R
R
.Product Quantity on Hand
X
R
R
RU
R
RU
Order
Ordered Product
Product
SS
SS
SS
SS
R
CRUD
R
CRUD
R
CRUD
R
CRUD
R
CRUD
R
CRUD
R
CRUD
R
CRUD
R
SS
SS
SS
SS
SS
ALL
R
CRUD
R
CRUD
CRUD
R
CRUD
RU
INDV = individual
ALL = ALL
SS = subset
X = no access
S = submit
C = create
R = read
U = update
25
CRUD
CRUD
D = delete
Network Modeling
System Concepts For Network Modeling

Synchronizing of System Models

Process and Interface Model Synchronization:
 The context diagram was previously introduced as an interface
model that documents how the system you are developing
interfaces to business, other systems, and other organizations.
 Data flow diagrams document the system’s process response to
various business and temporal events.
 Both models should be synchronized.
26
Network Modeling
System Concepts For Network Modeling

Synchronizing of System Models

Process and Network Model Synchronization:
 Process models illustrate the essential work to be performed by
the system as a whole.
 Network models identify the locations where work is to be
performed.
 Some work may be unique to one location. Other work may be
performed at multiple locations.
 Before designing the information system, what processes must
be performed at which locations should be identified and
documented.
27
Network Modeling
System Concepts For Network Modeling

Synchronizing of System Models

Process and Network Model Synchronization:
 Synchronization of the process and network models can be
accomplished through a Process-to-Location-Association
Matrix.
• A Process-to-Location-Association Matrix is a table in which the
rows indicate processes (event or elementary processes); the
columns indicate locations, and the cells (the intersection rows and
columns) document which processes must be performed at which
locations.
28
Process Customer Change to Outstanding Order
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
. Warehose
Process Customer Order Cancellation
X
San Diego
Process New Customer Order
X
. Sales
X
San Francisco
X
. Warehouse
Process Customer Change of Address
X
. Sales
Process Customer Credit Application
X
Boston
X
. Sales
Process Customer Application
. Accounts Receivable
. Warehouse
. Advertsing
. Marketing
Process
Customers
Kansas City
Network Modeling
X
X
Process Internal Customer Credit Chnage
Process Internal Change to Customer Order
Process New Product Addition
X
Process Product Withdrawal from Market
X
Process Product Price Chnage
X
Process Product Chnage to Product Specification
X
X
X
Process Product Inventory Adjustment
29
X
X
Network Modeling
The Process of Logical Network Modeling

Network Modeling during Strategic Systems Planning
Projects

Many systems planning methodologies and techniques result in a
network architecture to guide the design of all future computer
networks and applications that use those networks.
 May be a traditional map, or a top-down decomposition
diagram that logically groups locations.
 Association matrices are also typically used to provide an initial
mapping of data entities to locations, and processes to locations
30
Network Modeling
The Process of Logical Network Modeling

Network Modeling during Systems Analysis


During the study phase of systems analysis, a project team should
review any existing network models, logical or physical.
In the definition phase of systems analysis, network modeling
becomes more important.
 If a network model already exists, it is expanded or refined to
reflect new application requirements.
 If a network model does not exist, a network model should be
built from scratch.
31
Network Modeling
The Process of Logical Network Modeling

Looking Ahead to Systems Design


The logical application network model from systems analysis
describes business networking requirements, not technical
solutions.
In systems design, network models must become more technical –
they must become physical network models that will guide the
technical distribution and duplication of the other physical system
components, namely, DATA, PROCESSES, and INTERFACES.
32
Network Modeling
The Process of Logical Network Modeling

Fact-Finding and Information Gathering for Network
Modeling



Like all system models, network models are dependent on
appropriate facts and information as supplied by the user
community.
Facts can be collected by sampling of existing forms and files;
research of similar systems; surveys of users and management; and
interviews of users and management.
The fastest method of collecting facts and information, and
simultaneously constructing and verifying the process models is
Joint Application Development (JAD).
 JAD uses a carefully facilitated group meeting to collect the
facts, build the models, and verify the models – usually in one
or two full-day sessions.
33
Network Modeling
The Process of Logical Network Modeling

Computer-Aided Systems Engineering (CASE) for
Network Modeling


Network models should be stored in the repository.
Computer-aided systems engineering (CASE) technology provides
the repository for storing various models and their detailed
descriptions.
34
Network Modeling
How to Construct Logical Network Models

Location Decomposition Diagram


Decomposition diagrams are used to logically decompose and
group locations.
Building The Location Decomposition Diagram



First, brainstorm your locations. Think of all of the places where
direct and indirect users of your system will be located.
To group locations in the decomposition diagram, keep similar
locations on the same level or within the same branch of the tree.
Clustering reduces clutter through simplification; however, there is
a danger of oversimplifying the model.
 Cluster a location or its users if the data and processing
requirements for all users are expected to be the same.
35
Network Modeling
Member Services
Sys. Geography
Prospective
Members
Marketing
Office
Members
Advertising
Office
Indianapolis
Sales Office
Baltimore
Warehouse
Sales Office
Warehouse
Seattle
Portland
Sales Office
Warehouse
Sales
Managers
Sales
Managers
Sales
Managers
Order Entry
Clerks
Order Entry
Clerks
Order Entry
Clerks
36
Network Modeling
How to Construct Logical Network Models

Location Connectivity Diagram


Location connectivity diagrams can be initially drawn at a highlevel to communicate general information. More detailed
information can be added to subsequent diagrams.
In the following slides, the first location connectivity diagram
drawn is a systemwide model. It will include any external locations
and locations that have sublocations. The second diagram reveals
an exploded view.
37
Network Modeling
Baltimore
Prospective
Members
575 miles
indeterminate distance
Indianapolis
Seattle
2250 miles
Members
indeterminate distance
2250 miles
Portland
38
Network Modeling
Members (n)
Indeterminate Distance
Balt. Order Entry
Clerks (8)
< 20 feet
575 miles
Prospective
Members (n)
< 150 feet
Balt. Sales
Managers (2)
200 miles
Seattle Sales
Managers (2)
Indpls. Warehouse
Stations (3)
Indeterminate Distance
Indpls. Order Entry
Clerks (12)
Balt. Warehouse
Stations (3)
2800 miles
2250 miles
Port. Warehouse
Stations (2)
< 50 feet
575 miles
580 feet
< 40 feet
2250 miles
Members (n)
Indeterminate Distance
Indpls. Sales
Managers (3)
Members (n)
Indeterminate Distance
Seattle Order Entry
Clerks (7)
250 feet
Figure 7.10
Indpls. Advertising
Office (4)
39
50 feet
Indpls. Marketing
Office (3)
Network Modeling
The Next Generation

The Next Generation


The demand for logical network modeling skills will remain strong
so long as the trend towards distributed computing remains strong.
CASE tool support will evolve after methodologies, since CASE
tool engineers are reluctant to invest time and effort prior to some
semblance of a widely accepted methodological standard.
40
Network Modeling
Summary






Introduction
Network Modeling - Not Just For Computer
Networks
System Concepts For Network Modeling
The Process of Logical Network Modeling
How to Construct Logical Network Models
The Next Generation
41