1. Architectural Design Yonsei University 2 nd Semester, 2014 Sanghyun Park

2. Topics Covered  System structuring  Modular decomposition  Control models  Reference architectures

3. Software Architecture  Architectural design is the design process for identifying the __________ making up a system and the framework for sub-system _______ and communication  The output of this design process is a description of the software __________ sub-systems control architecture

4. Architectural Design  An ______ stage of the system design process  Represents the ____ between requirement specification and design processes  In many cases, carried out in ______ with some specification activities  Involves identifying major system components and their communications early link parallel

5. Architectural Design Process  System structuring  The system is decomposed into several principal __________ and communications between these sub-systems are identified  Modular decomposition  The identified sub-systems are decomposed into ________  Control modeling  A model of the ________ relationships between the different parts of the system is established sub-systems modules control

6. Sub-Systems And Modules  A sub-system is a system in its own right whose operation is __________ of the services provided by other sub-systems  A module is a system component that provides _______ to other components but would not normally be considered as an independent system independent services

7. Architectural Styles (1/3)  The system architecture affects the performance, robustness, and maintainability of a system  Therefore the particular style and structure chosen for an application may depend on the ____________ requirements non-functional

8. Architectural Styles (2/3)  Performance  _______ critical operations and ________ communications  Use ______ rather than fine-grain components  Security  Use a _______ architecture with critical assets in the inner layers Localizeminimize large layered

9. Architectural Styles (3/3)  Safety  _______ safety-critical operations in a small number of sub-systems  Availability  Include _________ components for fault tolerance  Maintainability  Use ____-grain, self-contained components Localize redundant fine

10. Architectural Conflicts  Using large-grain components improves __________ but reduces ____________  Introducing redundant data improves ________ but makes _______ more difficult  Localizing safety-related features usually means more ____________ so degraded ___________ performancemaintainability availability security communication performance

11. System Structuring (System Organization)  Concerned with decomposing the system into interacting sub-systems  The architectural design is normally expressed as a ______ diagram presenting an overview of the system structure  More specific models of the structure may be developed which show how systems share data, how they are distributed, and how they interface with each other block

12. Repository Model  Sub-systems must exchange data; this may be done in two ways:  Shared data is held in a central database or ________ and may be accessed by all sub-systems  Each sub-system maintains its ____ database and passes data explicitly to other sub-systems  When large amounts of data are to be shared, the _________ model of sharing is most commonly used repository own

13. CASE Toolset Architecture

14. Repository Model Characteristics (1/2)  Advantages  Efficient way to share large amount of data  Sub-systems which produce data need not be concerned with how that data is used by other sub- systems  Activities such as backup, security, access control and recovery from error are _________  Sharing model is visible through the repository ______ centralized schema

15. Repository Model Characteristics (2/2)  Disadvantages  Sub-systems must agree on a repository data model  Data ________ is difficult and expensive  No scope for specific management policies  Difficult to ________ the repository over a number of machines evolution distribute

16. Client-Server Model  Distributed system model which shows how data and processing are distributed across a range of components  Set of stand-alone ______ which provide specific services such as printing, data management, etc.  Set of ______ which call on these services  Network which allows clients to access servers servers clients

17. Film And Picture Library

18. Client-Server Characteristics  Advantages  Distribution of data is straightforward  Makes effective use of ________ systems. May require _______ hardware  Easy to add new servers or upgrade existing servers  Disadvantages  No shared data model so sub-systems use different data organization. Data __________ may be inefficient  _________ management in each server  No central _______ of names and services – it may be hard to find out what servers and services are available networked cheaper interchange Redundant register

19. Abstract Machine (Layered) Model  Used to model the interfacing of sub-systems  Organizes the system into a set of _____ each of which provides a set of services  Each layer defines an _______ machine whose machine language is used to implement the next level of abstract machine  When a layer interface changes, only the _______ layer is affected  Often difficult to structure systems in this way layers abstract adjacent

20. Version Management System Configuration management system layer Database system layer Operating system layer Object management system layer

21. Modular Decomposition  Another structural level where sub-systems are decomposed into ________  Two modular decomposition models covered  An _____ model where the system is decomposed into interacting objects  A function-oriented pipeline or ____-flow model where the system is decomposed into functional modules which ________ inputs to outputs modules object data transform

22. Object Models  Structure the system into a set of loosely coupled _______ with well-defined interfaces  Object-oriented decomposition is concerned with identifying object ______, their attributes, and operations  When implemented, objects are created from these classes and some control model is used to coordinate object operations objects classes

23. Invoice Processing System

24. Object Model Advantages  Objects are loosely coupled; so their implementation can be modified without affecting other objects  The objects may reflect ___-world entities  OO implementation languages are widely used  However, object ________ changes may cause problems and _______ entities may be hard to represent as objects real interface complex

25. Function-Oriented Pipelining (Data-Flow Models)  Functional transformations process their inputs to produce outputs  Variants of this approach are very common. When transformations are sequential with data processed in batches, this is a _____ sequential model  Not really suitable for _________ systems batch interactive

26. Invoice Processing Systems

27. Data-Flow Model Advantages  Supports transformation _____  Intuitive organization for stakeholder communication  Easy to add new transformation  Relatively simple to implement as either a concurrent or sequential system  However, requires a _______ format for data transfer along the pipeline and difficult to support _____-based interaction reuse common event

28. Control Models  Concerned with the control flow between sub-systems  Two general approaches can be identified  Centralized approach One sub-system has ______ responsibility for control, and starts and stops other sub-systems  Event-based approach Each sub-system can respond to events from other sub-systems or the system’s environment overall

29. Reference Architectures  Architectural models may be specific to some application _______  Two types of domain-specific model  Generic models which are abstractions from a number of _____ systems and which encapsulate the principal characteristics of these systems  Reference models which are more abstract, idealized model. Mainly used to communicate domain ________ and compare different architectures  Generic models are usually _________ models; Reference models are _________ models domain bottom-up top-down real concepts

30. Generic Models  Compiler model is a well-known example although other models exist in more specialized application domains  Lexical analyzer  Symbol table  Syntax analyzer  Syntax tree  Semantic analyzer  Code generator  Generic models may be reused directly in a design

31. Reference Models  Reference models are derived from a _____ of the application domain rather than from existing systems  May be used as a basis for system implementation or to compare different systems  Act as a standard against which systems can be evaluated  OSI model is a layered model for communication systems study

32. OSI Reference Model Presentation Session Transport Network Data link Physical 7 6 5 4 3 2 1 Communications medium Network Data link Physical Application Presentation Session Transport Network Data link Physical Application