Mapping Data Flow Models to the Palladio Component Model Stephan - PowerPoint PPT Presentation

Mapping Data Flow Models to the Palladio Component Model Stephan Seifermann, Dominik Werle, Mazen Ebada 06.11.2019 - Symposium on Software Performance 2019, Würzburg SOFTWARE DESIGN AND QUALITY GROUP, INSTITUTE FOR PROGRAM STRUCTURES AND DATA ORGANIZATION, KIT DEPARTMENT OF INFORMATICS www.kit.edu KIT – The Research University in the Helmholtz Association

Motivation Business WebUI Logic extract valid validate items order other other models tools Software Architects Sketched Architecture Architecture Models Analyses Are results Will quality goals be Are all models consistent? achieved? compatible? Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 2 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

High Level Challenges and Goals Challenges of multiple models and analyses • Consistency between separated models not trivial • Additional maintenance effort for multiple models • Analysis results have to be about same version Benefit of using dedicated models and analyses • Modeling approach tailored to quality analysis • Established analyses provide good predictions Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 3 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Topics in this Talk Covered Models • Data Flow Diagrams (DFDs) [DeMarco79] • Palladio Component Model (PCM) [Reussner16] Covered Challenges • Consistency: Mapping rules between DFDs and PCM • Maintenance Effort: Idea for automated PCM stub generation Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 4 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

State of the Art Relation Data Flow and Control Flow Resolving ambiguities Diagrams • Mapping DFD to object-oriented • Control flow nodes in DFDs design [Alabiso88] [WardMellor86] handling of ambiguities and breaks separation between design space not clear requirements and design • Mapping DFD to fine-grained • Relation inputs and outputs behavior [LarsenPlatToetenel94] [TsePong89][BrunzaWeide93] no realistic DFDs (1:1 mapping Does not match required of inputs to outputs) level of abstraction Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 5 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Foundations Data Flow Diagrams • Defined by DeMarco • Sources, sinks, processors, files, data flows Data Mapping Extension • Mapping of input data to output data [BrunzaWeide93] • 𝐹 : Set of all entities • 𝐸 : Set of all data • 𝑓, 𝐸 𝑗𝑜 , 𝐸 𝑝𝑣𝑢 : Mapping with e ∈ 𝐹 ⋀ 𝐸 𝑗𝑜 , 𝐸 𝑝𝑣𝑢 ∈ 𝐸 Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 6 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Running Example order Customer pay Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 7 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Running Example order order Customer pay Customer, ∅, order Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 8 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Running Example order order orderConfirmation bill Customer pay Customer, ∅, order order, order , orderConfirmation, bill Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 9 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Running Example order order orderConfirmation bill Customer creditCardData pay Customer, ∅, order Customer, orderConfirmation , creditCardData order, order , orderConfirmation, bill Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 10 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Running Example order order orderConfirmation bill paidBill Customer creditCardData pay payConfirmation Customer, ∅, order Customer, orderConfirmation , creditCardData order, order , orderConfirmation, bill pay, creditCardData,bill , paidBill, payConfirmation Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 11 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Coverage of Views by Mapping Rules Structural View covered • Components • System Behavioral View covered • Usage • Component Behavior Deployment not covered • Resources • Allocation Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 12 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Structural View – Components «DataType» «DataType» «DataType» «DataType» «DataType» Order CCD Bill OrderConfirmation PayConfirmation «Interface» Order IOrder «Interface» order(Order) : IPullBill OrderConfirmation getBill() : Bill «Interface» Pay IPay pay(CCD) : order, order , orderConfirmation, bill PayConfirmation pay, creditCardData,bill , paidBill, payConfirmation Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 13 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Structural View – System «System» OrderSystem «AssemblyContext» Order «AssemblyContext» Pay Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 14 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Behavioral View – Usage «UsageScenario» «Workload» Customer «ScenarioBehavior» Customer «ELSC» «Delay» «ELSC» call order wait call pay Customer, ∅, order Customer, orderConfirmation , creditCardData Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 15 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Behavioral View – Component Behavior «RDSEFF» Pay::pay «InternalAction» «ECA» process payment get bill Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 16 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization

Mapping Data Dependencies to Signatures Customer «Interface» order IOrder orderCon- firmation order(Order) : Bill, OrderConfirmation order bill Assumptions pay • Identification of caller • All data has to originate from and go to caller order, order , orderConfirmation, bill Motivation State of the Art Foundations Running Example Mappings Process Summary Stephan Seifermann, Dominik Werle, Mazen Ebada – 17 06 November 2019 Software Design and Quality Group Mapping Data Flow Models to the Palladio Component Model Institute for Program Structures and Data Organization