Object-Oriented Software Engineering

Conquering Complex and Changing Systems

Lecture 2
Chapter 6,
System Design
Preliminaries

♦ Guest lectures
w December 15: Andersen Consulting
w December 22: Viant
w January 18: sd&m

♦ Tomorrow December 1:
w Design Pattern I lecture by Tao Zhang

♦ For STARS students:

w SDD milestone moved to December 8
w SDD review moved to December 11

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 2
Exercise 6.1 from last week

6.1 Decomposing a system into subsystems reduces the

complexity developers have to deal with by simplifying the
parts and increasing their coherence. Decomposing a system
into simpler parts usually results into increasing a different kind
of complexity: Simpler parts also means a larger number of
parts and interfaces.
If coherence is the guiding principle driving developers to
decompose a system into small parts, which competing
principle drives them to keep the total number of parts small?

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 3
Solution to exercise 6.1

♦ Decreasing coupling is the principle that competes with

increasing coherence. A large number of parts will result in a
large number of interfaces and many dependencies among the
parts.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 4
Exercises (continued)

6.2 In Section 6.4.2 on page 193 in the book, we classified design

goals into five categories: performance, dependability, cost,
maintenance, and end user. Assign one or more categories to
each of the following goals:
w Users must be given a feedback within 1 second after they issue any
command.
w The TicketDistributor must be able to issue train tickets, even in the
event of a network failure.
w The housing of the TicketDistributor must allow for new buttons to
be installed in the event the number of different fares increases.
w The AutomatedTellerMachine must withstand dictionary attacks
(i.e., users attempting to discover a identification number by
systematic trial).
w The user interface of the system should prevent users from issuing
commands in the wrong order.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 5
Solution to exercise 6.2

♦ Users must be given a feedback within 1 second after they

issue any command. [Performance]
♦ The TicketDistributor must be able to issue train tickets, even
in the event of a network failure. [Dependability]
♦ The housing of the TicketDistributor must allow for new
buttons to be installed in the event the number of different fares
increases. [Maintenance]
♦ The AutomatedTellerMachine must withstand dictionary
attacks (i.e., users attempting to discover a identification
number by systematic trial). [Dependability]
♦ The user interface of the system should prevent users from
issuing commands in the wrong order. [End user]

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 6
Overview

System Design I (previous lecture)

0. Overview of System Design
1. Design Goals
2. Subsystem Decomposition

System Design II
3. Concurrency
4. Hardware/Software Mapping
5. Persistent Data Management
6. Global Resource Handling and Access Control
7. Software Control
8. Boundary Conditions

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 7
3. Concurrency

♦ Identify concurrent threads and address concurrency issues.

♦ Design goal: response time, performance.

♦ Threads
w A thread of control is a path through a set of state diagrams on
which a single object is active at a time.
w A thread remains within a state diagram until an object sends an
event to another object and waits for another event
w Thread splitting: Object does a nonblocking send of an event.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 8
Concurrency (continued)

♦ Two objects are inherently concurrent if they can receive

events at the same time without interacting

♦ Inherently concurrent objects should be assigned to different

threads of control

♦ Objects with mutual exclusive activity should be folded into a

single thread of control (Why?)

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 9
Concurrency Questions

♦ Which objects of the object model are independent?

♦ What kinds of threads of control are identifiable?
♦ Does the system provide access to multiple users?
♦ Can a single request to the system be decomposed into multiple
requests? Can these requests be handled in parallel?

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 10
Implementing Concurrency

♦ Concurrent systems can be implemented on any system that

provides
w physical concurrency (hardware)
or
w logical concurrency (software)

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 11
4. Hardware Software Mapping

♦ This activity addresses two questions:

w How shall we realize the subsystems: Hardware or Software?
w How is the object model mapped on the chosen hardware &
software?
t Mapping Objects onto Reality: Processor, Memory, Input/Output
t Mapping Associations onto Reality: Connectivity
♦ Much of the difficulty of designing a system comes from
meeting externally-imposed hardware and software constraints.
w Certain tasks have to be at specific locations

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 12
Mapping the Objects

♦ Processor issues:
w Is the computation rate too demanding for a single processor?
w Can we get a speedup by distributing tasks across several
processors?
w How many processors are required to maintain steady state load?
♦ Memory issues:
w Is there enough memory to buffer bursts of requests?
♦ I/O issues:
w Do you need an extra piece of hardware to handle the data
generation rate?
w Does the response time exceed the available communication
bandwidth between subsystems or a task and a piece of hardware?

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 13
Mapping the Subsystems Associations: Connectivity
♦ Describe the physical connectivity of the hardware
w Often the physical layer in ISO’s OSI Reference Model
t Which associations in the object model are mapped to physical
connections?
t Which of the client-supplier relationships in the analysis/design model
correspond to physical connections?

♦ Describe the logical connectivity (subsystem associations)

w Identify associations that do not directly map into physical
connections:
t How should these associations be implemented?

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 14
Connectivity in Distributed Systems
♦ If the architecture is distributed, we need to describe the network
architecture (communication subsystem) as well.
♦ Questions to ask
w What are the transmission media? (Ethernet, Wireless)
w What is the Quality of Service (QOS)? What kind of communication
protocols can be used?
w Should the interaction asynchronous, synchronous or blocking?
w What are the available bandwidth requirements between the
subsystems?
t Stock Price Change -> Broker
t Icy Road Detector -> ABS System

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 15
Typical Example of a Physical Connectivity Drawing
Application Application Application
Client Client Client

TCP/IP Ethernet
LAN

Communication
Agent for
Application Clients LAN Global
Data
Communication Server
Communication
Backbone Network Agent for Data
Agent for OODBMS
Server
Application Clients

Communication Global
Agent for Data Data
Server Server
LAN RDBMS

Local Data Global Data

Server Server

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 16
Hardware/Software Mapping Questions

♦ What is the connectivity among physical units?

w Tree, star, matrix, ring
♦ What is the appropriate communication protocol between the
subsystems?
w Function of required bandwidth, latency and desired reliability
♦ Is certain functionality already available in hardware?
♦ Do certain tasks require specific locations to control the
hardware or to permit concurrent operation?
w Often true for embedded systems
♦ General system performance question:
w What is the desired response time?

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 17
Drawing Subsystems in UML

♦ System design must model static and dynamic structures:

w Component Diagrams for static structures
t show the structure at design time or compilation time
w Deployment Diagram for dynamic structures
t show the structure of the run-time system

♦ Note the lifetime of components

w Some exist only at design time
w Others exist only until compile time
w Some exist at link or runtime

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 18
Component Diagram

♦ Component Diagram
w A graph of components connected by dependency relationships.
w Shows the dependencies among software components
t source code, linkable libraries, executables
♦ Dependencies are shown as dashed arrows from the client
component to the supplier component.
w The kinds of dependencies are implementation language specific.
♦ A component diagram may also be used to show dependencies
on a façade:
w Use dashed arrow the corresponding UML interface.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 19
Component Diagram Example

Scheduler reservations

UML Component
UML Interface

Planner update

GUI

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 20
Deployment Diagram

♦ Deployment diagrams are useful for showing a system design

after the following decisions are made
w Subsystem decomposition
w Concurrency
w Hardware/Software Mapping

♦ A deployment diagram is a graph of nodes connected by

communication associations.
w Nodes are shown as 3-D boxes.
w Nodes may contain component instances.
w Components may contain objects (indicating that the object is part
of the component)

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 21
Deployment Diagram Example
Compile Time
Dependency

:HostMachine
<<database>>
meetingsDB

:Scheduler

Runtime
Dependency

:PC

:Planner

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 22
5. Data Management
♦ Some objects in the models need to be persistent
w Provide clean separation points between subsystems with well-
defined interfaces.
♦ A persistent object can be realized with one of the following
w Data structure
t If the data can be volatile
w Files
t Cheap, simple, permanent storage
t Low level (Read, Write)
t Applications must add code to provide suitable level of abstraction
w Database
t Powerful, easy to port
t Supports multiple writers and readers

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 23
File or Database?

♦ When should you choose a file?

w Are the data voluminous (bit maps)?
w Do you have lots of raw data (core dump, event trace)?
w Do you need to keep the data only for a short time?
w Is the information density low (archival files,history logs)?
♦ When should you choose a database?
w Do the data require access at fine levels of details by multiple users?
w Must the data be ported across multiple platforms (heterogeneous
systems)?
w Do multiple application programs access the data?
w Does the data management require a lot of infrastructure?

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 24
Database Management System

♦ Contains mechanisms for describing data, managing persistent

storage and for providing a backup mechanism
♦ Provides concurrent access to the stored data
♦ Contains information about the data (“meta-data”), also called
data schema.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 25
Issues To Consider When Selecting a Database
♦ Storage space
w Database require about triple the storage space of actual data
♦ Response time
w Mode databases are I/O or communication bound (distributed
databases). Response time is also affected by CPU time, locking
contention and delays from frequent screen displays
♦ Locking modes
w Pessimistic locking: Lock before accessing object and release when
object access is complete
w Optimistic locking: Reads and writes may freely occur (high
concurrency!) When activity has been completed, database checks if
contention has occurred. If yes, all work has been lost.
♦ Administration
w Large databases require specially trained support staff to set up
security policies, manage the disk space, prepare backups, monitor
performance, adjust tuning.
Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 26
Object-Oriented Databases

♦ Support all fundamental object modeling concepts

w Classes, Attributes, Methods, Associations, Inheritance
♦ Mapping an object model to an OO-database
w Determine which objects are persistent.
w Perform normal requirement analysis and object design
w Create single attribute indices to reduce performance bottlenecks
w Do the mapping (specific to commercially available product).
Example:
t In ObjectStore, implement classes and associations by preparing C++
declarations for each class and each association in the object model

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 27
Relational Databases

♦ Based on relational algebra

♦ Data is presented as 2-dimensional tables. Tables have a
specific number of columns and and arbitrary numbers of rows
w Primary key: Combination of attributes that uniquely identify a
row in a table. Each table should have only one primary key
w Foreign key: Reference to a primary key in another table
♦ SQL is the standard language defining and manipulating tables.
♦ Leading commercial databases support constraints.
w Referential integrity, for example, means that references to entries
in other tables actually exist.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 28
Mapping an object model to a relational database

♦ UML object models can be mapped to relational databases:

w Some degradation occurs because all UML constructs must be
mapped to a single relational database construct - the table.
♦ UML mappings
w Each class is mapped to a table
w Each class attribute is mapped onto a column in the table
w An instance of a class represents a row in the table
w A many-to-many association is mapped into its own table
w A one-to-many association is implemented as buried foreign key
♦ Methods are not mapped

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 29
Turning Object Models into Tables I
Many-to-Many Associations: Separate Table for Association

City * Serves Airport

*
airportCode
cityName airportName

Separate
Primary Key Table

City Table Airport Table Serves Table

airportCode airportName cityName airportCode

cityName IAH Intercontinental Houston IAH
Houston HOU Hobby Houston HOU
Albany ALB Albany County Albany ALB
Munich MUC Munich Airport Munich MUC
Hamburg HAM Hamburg Airport Hamburg HAM

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 30
Turning Object Models into Tables II

1-To-Many or Many-to-1 Associations: Buried Foreign Keys

Portfolio
Transaction *
portfolioID
transactionID ...
Foreign Key

Transaction Table Portfolio Table

transactionID portfolioID portfolioID ...

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 31
Data Management Questions
♦ Should the data be distributed?
♦ Should the database be extensible?
♦ How often is the database accessed?
♦ What is the expected request (query) rate? In the worst case?
♦ What is the size of typical and worst case requests?
♦ Do the data need to be archived?
♦ Does the system design try to hide the location of the databases
(location transparency)?
♦ Is there a need for a single interface to access the data?
♦ What is the query format?
♦ Should the database be relational or object-oriented?

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 32
6. Global Resource Handling

♦ Discusses access control

♦ Describes access rights for different classes of actors
♦ Describes how object guard against unauthorized access

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 33
Global Resource Questions

♦ Does the system need authentication?

♦ If yes, what is the authentication scheme?
w User name and password? Access control list
w Tickets? Capability-based
♦ What is the user interface for authentication?
♦ Does the system need a network-wide name server?
♦ How is a service known to the rest of the system?
w At runtime? At compile time?
w By Port?
w By Name?

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 34
7. Decide on Software Control

A. Choose implicit control (non-procedural or declarative

languages)
w Rule-based systems
w Logic programming
B. Or choose explicit control (procedural languages)
w Centralized control
t 1. Procedure-driven control
– Control resides within program code. Example: Main program
calling procedures of subsystems.
– Simple, easy to build

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 35
Software Control (continued)

t 2. Event-driven control
– Control resides within a dispatcher who calls subsystem functions
via callbacks.
– Flexible, good for user interfaces
w Decentralized control
t Control resided in several independent objects (supported by some
languages).
t Possible speedup by parallelization, increased communication
overhead.
t Example: Message based system.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 36
Procedure-Driven Control Example

op1()
module1 module2

op3() op2()
module3

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 37
Event-Based System Example: MVC

♦ Smalltalk-80 Model-View-Controller
♦ Client/Server Architecture

:Control

Update :View
Model has changed
Update Update

:Model :View
:View

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 38
Centralized vs. Decentralized Designs

♦ Should you use a centralized or decentralized design?

♦ Centralized Design
w One control object or subsystem ("spider") controls everything
w Change in the control structure is very easy
w Possible performance bottleneck
♦ Decentralized Design
w Control is distributed
w Spreads out responsibility
w Fits nicely into object-oriented development

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 39
8. Boundary Conditions

♦ Most of the system design effort is concerned with steady-state

behavior.
♦ However, the system design phase must also address the
initiation and finalization of the system.
w Initialization
t Describes how the system is brought from an non initialized state to
steady-state ("startup use cases”).
w Termination
t Describes what resources are cleaned up and which systems are
notified upon termination ("termination use cases").
w Failure
t Many possible causes: Bugs, errors, external problems (power supply).
t Good system design foresees fatal failures (“failure use cases”).

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 40
Boundary Condition Questions
♦ 8.1 Initialization
w How does the system start up?
t What data need to be accessed at startup time?

t What services have to registered?

w What does the user interface do at start up time?

t How does it present itself to the user?

♦ 8.2 Termination
w Are single subsystems allowed to terminate?
w Are other subsystems notified if a single subsystem terminates?
w How are local updates communicated to the database?
♦ 8.3 Failure
w How does the system behave when a node or communication link fails?
Are there backup communication links?
w How does the system recover from failure? Is this different from
initialization?
Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 41
Summary

In this lecture, we reviewed the activities of system design :

♦ Concurrency identification
♦ Hardware/Software mapping
♦ Persistent data management
♦ Global resource handling
♦ Software control selection
♦ Boundary conditions

Each of these activities revises the subsystem decomposition to

address a specific issue. Once these activities are completed,
the interface of the subsystems can be defined.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 42
Exercises

6.4 Consider a legacy, fax-based, problem-reporting system for

an aircraft manufacturer. You are part of a reengineering
project replacing the core of the system by a computer-based
system, which includes a database and a notification system.
The client requires the fax to remain an entry point for
problem reports. You propose an E-mail entry point.

Describe a subsystem decomposition, and possibly a design

pattern, which would allow both interfaces.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 43
Exercises (cont’d)

6.5 You are designing the access control policies for a Web-
based retail store:
w Customers access the store via the Web, browse product
information, input their address and payment information, and
purchase products.
w Suppliers can add new products, update product information,
and receive orders.
w The store owner sets the retail prices, makes tailored offers to
customers based on their purchasing profiles, and provides
marketing services.
You have to deal with three actors: StoreAdministrator,
Supplier, and Customer. Design an access control policy for
all three actors. Customers can be created via the Web,
whereas Suppliers are created by the StoreAdministrator.

Bernd Bruegge & Allen Dutoit Object-Oriented Software Engineering: Conquering Complex and Changing Systems 44