April 10, 2014CS410 – Software Engineering Lecture #16: System Design 1Design “There are two ways of constructing a software design: One way is to make it so simple that there are obviously no deficiencies, and the other way is to make it so complicated that there are no obvious deficiencies.” - C.A.R. Hoare

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 2 Why is Design so Difficult? Analysis: Focuses on the application domain Design: Focuses on the implementation domain Design knowledge is a moving targetDesign knowledge is a moving target The reasons for design decisions are changing very rapidlyThe reasons for design decisions are changing very rapidly  Halftime knowledge in software engineering: About 3-5 years  Cost of hardware rapidly sinking “Design window”: Time in which design decisions have to be madeTime in which design decisions have to be made

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 3 The Purpose of System Design Bridging the gap between desired and existing system in a manageable wayBridging the gap between desired and existing system in a manageable way Use Divide and ConquerUse Divide and Conquer We model the new system to be developed as a set of subsystemsWe model the new system to be developed as a set of subsystems Problem Existing System New System

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 4 System Design 1. Design Goals Definition Trade-offfs 5. Data Data structure Persistent Objects Files Databases Management Access control Security 6. Global Resource Handling 2. System Layers/Partitions Coherence/Coupling Decomposition 4. Hardware/ Special purpose Software Buy or Build Trade-off Allocation Connectivity Mapping 8. Boundary Conditions Initialization Termination Failure 3. Concurrency Identification of Threads 7. Software Control Monolithic Event-Driven Threads Conc. Processes

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 5 How to use the results from the Requirements Analysis for System Design Nonfunctional requirements:  Activity 1: Design Goals Definition Use Case model:  Activity 2: System decomposition (Selection of subsystems based on functional requirements, coherence, and coupling)  Activity 2: System decomposition (Selection of subsystems based on functional requirements, coherence, and coupling) Object model:  Activity 4: Hardware/software mapping  Activity 4: Hardware/software mapping  Activity 5: Persistent data management  Activity 5: Persistent data management

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 6 How to use the results from the Requirements Analysis for System Design Dynamic model:  Activity 3: Concurrency  Activity 3: Concurrency  Activity 6: Global resource handling  Activity 6: Global resource handling  Activity 7: Software control  Activity 7: Software control  Activity 8: Boundary conditions  Activity 8: Boundary conditions

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 7 Design Goals ReliabilityReliability ModifiabilityModifiability MaintainabilityMaintainability UnderstandabilityUnderstandability AdaptabilityAdaptability ReusabilityReusability EfficiencyEfficiency PortabilityPortability Traceability of requirementsTraceability of requirements Fault toleranceFault tolerance Backward-compatibilityBackward-compatibility Cost-effectivenessCost-effectiveness RobustnessRobustness High-performanceHigh-performance Good documentation Well-defined interfaces User-friendliness Reuse of components Rapid development Minimum # of errors Readability Ease of learning Ease of remembering Ease of use Increased productivity Low-cost Flexibility

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 8 Relationship Between Design Goals Reliability Low cost Increased productivity Backward-compatibility Traceability of requirements Rapid development Flexibility End User Portability Good documentation Runtime efficiency Client (Customer, Sponsor) Developer/ Maintainer Minimum # of errors Modifiability, Readability Reusability, Adaptability Well-defined interfaces Functionality User-friendliness Ease of use Ease of learning Fault tolerance Robustness

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 9 Typical Design Trade-offs Functionality vs. usabilityFunctionality vs. usability Cost vs. robustnessCost vs. robustness Efficiency vs. portabilityEfficiency vs. portability Rapid development vs. functionalityRapid development vs. functionality Cost vs. reusabilityCost vs. reusability Backward compatibility vs. readabilityBackward compatibility vs. readability

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 10 System Decomposition Subsystem (UML: Package) Collection of classes, associations, operations, events and constraints that are interrelatedCollection of classes, associations, operations, events and constraints that are interrelated Seed for subsystems: UML Objects and Classes.Seed for subsystems: UML Objects and Classes.Service: Group of operations provided by the subsystemGroup of operations provided by the subsystem Seed for services: Subsystem use casesSeed for services: Subsystem use cases Service is specified by Subsystem interface: Specifies interaction and information flow from/to subsystem boundaries, but not inside the subsystem.Specifies interaction and information flow from/to subsystem boundaries, but not inside the subsystem. Should be well-defined and small.Should be well-defined and small. Often called API: Application programmer’s interface, but this term should be used during implementation, not during System DesignOften called API: Application programmer’s interface, but this term should be used during implementation, not during System Design

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 11 Choosing Subsystems Criteria for subsystem selection: Most of the interaction should be within subsystems, rather than across subsystem boundaries (High coherence). Does one subsystem always call the other for the service?Does one subsystem always call the other for the service? Which of the subsystems call each other for service?Which of the subsystems call each other for service? Primary Question: What kind of service is provided by the subsystems (subsystem interface)?What kind of service is provided by the subsystems (subsystem interface)? Secondary Question: Can the subsystems be hierarchically ordered (layers)? What kind of model is good for describing layers and partitions?Can the subsystems be hierarchically ordered (layers)? What kind of model is good for describing layers and partitions?

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 12 Coupling and Coherence Goal: Reduction of complexity Coherence measures the dependence among classes in a subsystem. High coherence: The classes in the subsystem perform similar tasks and are related to each other (via associations)High coherence: The classes in the subsystem perform similar tasks and are related to each other (via associations) Low coherence: Lots of miscellaneous and auxiliary objects, no associationsLow coherence: Lots of miscellaneous and auxiliary objects, no associations

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 13 Coupling and Coherence Coupling measures dependencies between two subsystems. High coupling: Modifications to one subsystem will have high impact on the other subsystem (change of model, massive recompilation, etc.).High coupling: Modifications to one subsystem will have high impact on the other subsystem (change of model, massive recompilation, etc.). Low coupling: The two systems are relatively independent from each other.Low coupling: The two systems are relatively independent from each other. Subsystems should have as much coherence and as little coupling as possible. How can we achieve loose coupling?How can we achieve loose coupling? Which subsystems are highly coupled?Which subsystems are highly coupled?

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 14 Partitions and Layers A large system is usually decomposed into subsystems using both layers and partitions. Partitions vertically divide a system into several independent (or weakly-coupled) subsystems that provide services on the same level of abstraction. A layer is a subsystem that provides services to a higher level of abstraction A layer can only depend on lower layersA layer can only depend on lower layers A layer has no knowledge of higher layersA layer has no knowledge of higher layers

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 15 B:Subsystem A: Subsystem C:Subsystem E:Subsystem F:Subsystem G:SubsystemD:Subsystem Layer 1 Layer 2 Layer 3 Subsystem Decomposition into Layers Subsystem decomposition heuristics: No more than 7+/-2 subsystems at any levelNo more than 7+/-2 subsystems at any level More subsystems increase coherence but also complexity (more services)More subsystems increase coherence but also complexity (more services) No more than 5+/-2 layersNo more than 5+/-2 layers

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 16 Layer and Partition Relationships between Subsystems Layer relationship: Layer A “calls” Layer B (runtime).Layer A “calls” Layer B (runtime). Layer A “depends on” Layer B (“make” dependency, compile time).Layer A “depends on” Layer B (“make” dependency, compile time). Partition relationship: The subsystems have mutual but not deep knowledge about each other,The subsystems have mutual but not deep knowledge about each other, Partition A “calls” partition B and partition B “calls” partition APartition A “calls” partition B and partition B “calls” partition A

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 17 Virtual Machine (Dijkstra, 1965) A system should be developed by an ordered set of virtual machines, each built in terms of the ones below it. Problem Existing System Class1 attributes operations Class2attributes operations VM1 Class3attributes operations Class5attributes operations VM2 Class4attributes operationsClass6attributes operations VM3 Class7attributes operationsClass8attributes operations VM4

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 18 Virtual Machine A virtual machine is an abstraction that provides a set of attributes and operations.A virtual machine is an abstraction that provides a set of attributes and operations. A virtual machine is a subsystem connected to higher and lower level virtual machines by "provides services for" associations.A virtual machine is a subsystem connected to higher and lower level virtual machines by "provides services for" associations. Virtual machines can implement two types of software architecture: closed and open architectures.Virtual machines can implement two types of software architecture: closed and open architectures.

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 19 Closed Architecture (Opaque Layering) A virtual machine can only call operations from the layer directly below it.A virtual machine can only call operations from the layer directly below it. Design goal: High maintainabilityDesign goal: High maintainability Class1 attributes operations Class2attributes operations Class3attributes operations Class5attributes operationsClass4attributes operationsClass6attributes operations Class7attributes operationsClass8attributes operations VM1 VM2 VM3 VM4

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 20 Physical Frame Packet Bit ConnectionFormat Message Level of abstraction Application Presentation Session TransportNetwork DataLink Example of a Closed Architecture ISO’s OSI Reference Model ISO = International Standard OrganizationISO = International Standard Organization OSI = Open System InterconnectionOSI = Open System Interconnection Reference model defines 7 layers of network protocols and strict methods of communication between the layers.

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 21 Open Architecture (Transparent Layering) A virtual machine can call operations from any layer below it.A virtual machine can call operations from any layer below it. Design goal: Runtime efficiencyDesign goal: Runtime efficiency Class1 attributes operations Class2attributes operations Class3attributes operations Class5attributes operationsClass4attributes operationsClass6attributes operations Class7attributes operationsClass8attributes operations VM1 VM2 VM3 VM4

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 22 Properties of Layered Systems Layered systems are hierarchical. They are desirable because hierarchy reduces complexity.Layered systems are hierarchical. They are desirable because hierarchy reduces complexity. Closed architectures are more portable.Closed architectures are more portable. Open architectures are more efficient.Open architectures are more efficient. If a subsystem is a layer, it is often called a virtual machine.If a subsystem is a layer, it is often called a virtual machine.

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 23 Software Architectures Subsystem decomposition: Identification of subsystems, services, and their relationship to each other.Identification of subsystems, services, and their relationship to each other. Specification of the system decomposition is critical. Patterns for software architecture: Client/Server ArchitectureClient/Server Architecture Peer-To-Peer ArchitecturePeer-To-Peer Architecture Repository ArchitectureRepository Architecture Model/View/ControllerModel/View/Controller

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 24 Client/Server Architecture One or many servers provide services to instances of subsystems, called clients. One or many servers provide services to instances of subsystems, called clients. Client calls on the server, which performs some service and returns the result Client calls on the server, which performs some service and returns the result Client knows the interface of the server (its service) Client knows the interface of the server (its service) Server does not need to know the interface of the client Server does not need to know the interface of the client Response in general immediately Response in general immediately Users interact only with the client Users interact only with the client Client Server service1() service2() serviceN() … * * requesterprovider

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 25 Client/Server Architecture Often used in database systems: Front-end: User application (client)Front-end: User application (client) Back end: Database access and manipulation (server)Back end: Database access and manipulation (server) Functions performed by client: Customized user interfaceCustomized user interface Front-end processing of dataFront-end processing of data Initiation of server remote procedure callsInitiation of server remote procedure calls Access to database server across the networkAccess to database server across the network

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 26 Client/Server Architecture Functions performed by the database server: Centralized data managementCentralized data management Data integrity and database consistencyData integrity and database consistency Database securityDatabase security Concurrent operations (multiple user access)Concurrent operations (multiple user access) Centralized processing (for example archiving)Centralized processing (for example archiving)

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 27 Design Goals for Client/Server Systems Portability: Server can be installed on a variety of machines and operating systems and functions in a variety of networking environmentsServer can be installed on a variety of machines and operating systems and functions in a variety of networking environmentsTransparency: The server might itself be distributed, but should provide a single "logical" service to the userThe server might itself be distributed, but should provide a single "logical" service to the userPerformance: Client should be customized for interactive display-intensive tasksClient should be customized for interactive display-intensive tasks Server should provide CPU-intensive operationsServer should provide CPU-intensive operations

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 28 Design Goals for Client/Server Systems Scalability: Server has spare capacity to handle larger number of clientsServer has spare capacity to handle larger number of clientsFlexibility: Should be usable for a variety of user interfacesShould be usable for a variety of user interfacesReliability: System should survive individual node and/or communication link problemsSystem should survive individual node and/or communication link problems

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 29 Problems with Client/Server Architectures Client/server systems do not provide peer-to-peer communicationClient/server systems do not provide peer-to-peer communication Peer-to-peer communication is often neededPeer-to-peer communication is often needed Example: Database receives queries from application but also sends notifications to application when data have changedExample: Database receives queries from application but also sends notifications to application when data have changed

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 30 Peer-to-Peer Architecture Generalization of Client/Server ArchitectureGeneralization of Client/Server Architecture Clients can be servers and servers can be clientsClients can be servers and servers can be clients More difficult because of possibility of deadlocksMore difficult because of possibility of deadlocks Peer service1() service2() serviceN() … requester provider * * application1:DBUser database:DBMS application2:DBUser 1. updateData 2. changeNotification

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 31 Repository Architecture Subsystems access and modify data from a single data structureSubsystems access and modify data from a single data structure Subsystems are loosely coupled (interact only through the repository)Subsystems are loosely coupled (interact only through the repository) Control flow is dictated by central repository (triggers) or by the subsystems (locks, synchronization primitives)Control flow is dictated by central repository (triggers) or by the subsystems (locks, synchronization primitives) Subsystem Repository createData() setData() getData() searchData()

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 32 Examples of Repository Architecture Hearsay II speech understanding system (“Blackboard architecture”) Database Management Systems Modern Compilers LexicalAnalyzer SyntacticAnalyzer SemanticAnalyzer CodeGenerator Compiler SyntacticEditor ParseTree SymbolTable Repository SourceLevelDebugger Optimizer

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 33 Model/View/Controller Architecture Subsystems are classified into 3 different types Model subsystem: Responsible for application domain knowledgeModel subsystem: Responsible for application domain knowledge View subsystem: Responsible for displaying application domain objects to the userView subsystem: Responsible for displaying application domain objects to the user Controller subsystem: Responsible for sequence of interactions with the user and notifying views of changes in the model.Controller subsystem: Responsible for sequence of interactions with the user and notifying views of changes in the model.

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 34 Model/View/Controller Architecture MVC is a special case of a repository architecture: Model subsystem implements the central data structure; the Controller subsystem explicitly dictates the control flowModel subsystem implements the central data structure; the Controller subsystem explicitly dictates the control flow Controller Model subscriber notifier initiator * repository1 1 * View

April 10, 2014CS410 – Software Engineering Lecture #16: System Design 35 Sequence of Events :Controller:InfoView:Model:FolderView 1. Views subscribe to event 2.User types new filename 3. Request name change in model 4. Notify subscribers 5. Update views