SlideShare une entreprise Scribd logo
1  sur  23
Télécharger pour lire hors ligne
©paluno
Coordinated Run-time Adaptation of Variability-intensive Systems
An Application in Cloud Computing
Andreas Metzger†, Andreas Bayer†, Daniel Doyle*, Amir Molzam Sharifloo†, Klaus Pohl†, Florian Wessling†
† paluno (The Ruhr Institute for Software Technology), University of Duisburg Essen, Germany
* Intel, Ireland
©paluno
2
Motivation
& Problem
Statement
FCORE
Approach
FCORE
Models
FCORE
Analysis
Con-
clusion &
Outlook
©paluno
Motivation
Adaptive Software Systems
 Adaptive software systems can modify own structure and
behavior at run time to cope with dynamic changes in…
 M = machine (software)  self-healing
 W = world (context)  context-aware
 R = changing requirements  ???
 @runtime: adaptive systems monitor changes in M and W
or even R directly
 M, W |≠ R (requirements violation)  self-modification
3
M, W |= R ?
©paluno
Motivation
Coordination among Adaptive Systems
 Distributed systems (e.g., cloud systems or cyber-physical systems)
orchestrate many adaptive sub-systems
 Each sub-system may perform adaptations simultaneously and
independent of each other
 However, adaptations may affect shared phenomena, thus:
 Conflicts between adaptations may occur
 Synergies among adaptations may be missed
4
Shared
Phenomenon
M, W |= R ?
M, W |= R ?
M, W |= R ?
©paluno
Motivation
Use Case: Conflicts in Adaptive Cloud Systems
 „AdvancedTV“ Use Case from EU Project Cloud Wave
 Cloud application that offers services in parallel to running TV programme
 Two adaptive systems:
 Cloud Infrastructure (IaaS: CPU, RAM, …)
 Cloud Application (SaaS)
 Adaptations of Cloud Infrastructure
 Horizontal Scaling
 E.g., turning off virtual machines to save energy
 Vertical Scaling
 …
 Adaptations of Cloud Application
 Different levels of social media features
 No
 Partial
 Unlimited
 …
5
Performance
-- VM  -- Performance
++ Socia Media  -- Performance
©paluno
6
Motivation
& Problem
Statement
FCORE
Approach
FCORE
Models
FCORE
Analysis
Con-
clusion &
Outlook
©paluno
FCORE Approach
Main Ideas and Challenges
 Explicitly model adaptations and dependencies among systems
during design time
 Challenge 1: Developers must model adaptations of their systems
 Sufficiently compact, yet expressive modeling technique
 Challenge 2: Systems developed by different developers/organizations
 Suitably (small) common denominator to describe dependencies
among systems
 Analyze models at run time to determine conflicts and identify
optimizations (synergies)
 Challenge 3: Self-adaptation at run time must be fast enough to be
effective (otherwise may be too late)
 Efficient model analysis during system execution
7
©paluno
8
Motivation
& Problem
Statement
FCORE
Approach
FCORE
Models
FCORE
Analysis
Con-
clusion &
Outlook
©paluno
FCORE Models
Main Underlying Concepts
 FCORE = DSPL Feature Model + Goal Models
9
Concepts from
Dynamic Software
Product Lines
Feature Models to
describe
adaptations
Concepts from
Goal Models to
describe
dependencies
via shared
phenomena
Main underlying assumption:
“Known Unkowns!“
©paluno
FCORE Models
Why DSPL Models? (Challenge 1)
 DSPLs can build on proven engineering foundations of SPLs!
 DSPL extend existing software product line engineering approaches
by moving their capabilities to run time
 Variability binding is postponed to run time, allowing a DSPL to
activate or deactivate certain features
 Configurations of a DSPL are expressed in terms of a product line
variability model, usually a feature model
10
Classical SPL Dynamic SPL
variability describes
different pos-
sible software systems
variability describes
different possible
configurations (i.e.,
adaptations) of the same
system
©paluno
FCORE Models
Which Kinds of DSPL Models? (Challenges 1&3)
Approach Expressiveness Analysis
Basic-FM
High redundancy in
models (replication of FM
sub-trees)
Cardinalities only 1..1 /
0..n
SAT solver
Cardinality-Based
FM
Alternative-Groups
Cardinalities n..m
Feature-Cardinalities n..m
(i.e., instantiation of
features)
SAT solver
Extended-FM
Feature-Attributes
(Integer, Enumeration, …)
CSP solver
11
©paluno
FCORE Models
Why Goal Models? (Challenge 2)
 Soft Goals provide high-level of abstraction to describe
influences of features (~ “tasks”) on goal satisfaction
 Well-known from requirements engineering
 Defining dependencies among systems requires
agreeing on a set of shared soft goals
 E.g., in cloud computing, these soft goals may be derived from
standardized QoS catalogues for SLAs
12
©paluno
FCORE Models
CloudWave Use Case: Simplified
13
©paluno
FCORE Models
CloudWave Use Case: SaaS
14
©paluno
FCORE Models
CloudWave Use Case: IaaS
15
©paluno
16
Motivation
& Problem
Statement
FCORE
Approach
FCORE
Models
FCORE
Analysis
Con-
clusion &
Outlook
©paluno
FCORE Analysis
Main Underlying Strategy (Challenge 3)
 Formalize FCORE Model as CSP (justification see above)
 Perform automated reasoning on formalization
 Two main usages:
 FCORE Filter: Validity check of given configurations
(= detecting conflicts)
 E.g., 1 CPU + Unlimited Social Media  violation of high performance
 FCORE Search: Search for configurations with high goal satisfaction
(= exploiting synergies)
 E.g., 6 CPUs + No Social Media  high performance + low costs
17
©paluno
FCORE Analysis
Formalization: Features
18
A
= Feature
selected
 Feature
 Requires-Relation
 Excludes-Relation
 Feature Group
©paluno
FCORE Analysis
Formalization: Goals
20
 Softgoals and Attributes
Softgoal satisfaction:
sgVal = [-1.0, +1.0]
©paluno
FCORE Analysis
Performance (Challenge 3)
 FCORE Filter
 No performance issues
 Just compute goal satisfaction for given configuration
 Ca. 2ms for cloud use case
 FCORE Search
 CSP to find optimal configurations (maximize sgVal)
 Experimental results for cloud use case
21
©paluno
22
Motivation
& Problem
Statement
FCORE
Approach
FCORE
Models
FCORE
Analysis
Con-
clusion &
Outlook
©paluno
Conclusion and Outlook
 Concluded: FCORE as an approach for coordinating among
adaptive, variability intensive systems
 Building on DSPLs
 Offering Modelling + Analysis
 Exemplified for the case of cloud computing
 Ongoing:
 Implementation as part of CloudWave Adaptation Engine
(jointly with IBM and intel)
 Future:
 Handling “Unknown Unknowns”:
Extending DSPLs with
dynamic learning
and evolution
23
©paluno
The research leading to these results has
received funding from the European Union's
Seventh Framework Programme FP7/2007-
2013 under grant agreement 610802
(CloudWave)
http://www.cloudwave-fp7.eu/
Thank You!

Contenu connexe

En vedette (20)

BBBSN 101
BBBSN 101BBBSN 101
BBBSN 101
 
000739
000739000739
000739
 
PD Stone ISDN Article - Magazine
PD Stone ISDN Article - MagazinePD Stone ISDN Article - Magazine
PD Stone ISDN Article - Magazine
 
Cursos de Férias
Cursos de FériasCursos de Férias
Cursos de Férias
 
000691
000691000691
000691
 
NewsWhip Syndication
NewsWhip SyndicationNewsWhip Syndication
NewsWhip Syndication
 
000716
000716000716
000716
 
21 de mayo
21 de mayo21 de mayo
21 de mayo
 
000723
000723000723
000723
 
Final report
Final reportFinal report
Final report
 
Coverletter
CoverletterCoverletter
Coverletter
 
000930
000930000930
000930
 
000686
000686000686
000686
 
000679
000679000679
000679
 
000658
000658000658
000658
 
Atención
AtenciónAtención
Atención
 
000718
000718000718
000718
 
000694
000694000694
000694
 
Aman goel's PPT
Aman goel's PPTAman goel's PPT
Aman goel's PPT
 
000670
000670000670
000670
 

Similaire à Coordinated run-time adaptation of variability-intensive systems: an application in cloud computing (VACE 2016)

Conditional Execution - A Pattern for the Implementation of Fine-Grained Vari...
Conditional Execution - A Pattern for the Implementation of Fine-Grained Vari...Conditional Execution - A Pattern for the Implementation of Fine-Grained Vari...
Conditional Execution - A Pattern for the Implementation of Fine-Grained Vari...
Jadson Santos
 
Software engineering Questions and Answers
Software engineering Questions and AnswersSoftware engineering Questions and Answers
Software engineering Questions and Answers
Bala Ganesh
 
Ch16-Software Engineering 9
Ch16-Software Engineering 9Ch16-Software Engineering 9
Ch16-Software Engineering 9
Ian Sommerville
 
term paper for cbd models
term paper for cbd modelsterm paper for cbd models
term paper for cbd models
Sukhdeep Singh
 
Learning Software Performance Models for Dynamic and Uncertain Environments
Learning Software Performance Models for Dynamic and Uncertain EnvironmentsLearning Software Performance Models for Dynamic and Uncertain Environments
Learning Software Performance Models for Dynamic and Uncertain Environments
Pooyan Jamshidi
 
CS587 Project - Raychaudhury,Shaalmali
CS587 Project - Raychaudhury,ShaalmaliCS587 Project - Raychaudhury,Shaalmali
CS587 Project - Raychaudhury,Shaalmali
sagar.247
 

Similaire à Coordinated run-time adaptation of variability-intensive systems: an application in cloud computing (VACE 2016) (20)

MODEL DRIVEN ARCHITECTURE, CONTROL SYSTEMS AND ECLIPSE
MODEL DRIVEN ARCHITECTURE, CONTROL SYSTEMS AND ECLIPSEMODEL DRIVEN ARCHITECTURE, CONTROL SYSTEMS AND ECLIPSE
MODEL DRIVEN ARCHITECTURE, CONTROL SYSTEMS AND ECLIPSE
 
Dynamic Component Deployment and (Re) Configuration Using a Unified Framework
Dynamic Component Deployment and (Re) Configuration Using a Unified FrameworkDynamic Component Deployment and (Re) Configuration Using a Unified Framework
Dynamic Component Deployment and (Re) Configuration Using a Unified Framework
 
Object Orientation Fundamentals
Object Orientation FundamentalsObject Orientation Fundamentals
Object Orientation Fundamentals
 
Model-driven Framework for Dynamic Deployment and Reconfiguration of Componen...
Model-driven Framework for Dynamic Deployment and Reconfiguration of Componen...Model-driven Framework for Dynamic Deployment and Reconfiguration of Componen...
Model-driven Framework for Dynamic Deployment and Reconfiguration of Componen...
 
DesignPrinciples-and-DesignPatterns
DesignPrinciples-and-DesignPatternsDesignPrinciples-and-DesignPatterns
DesignPrinciples-and-DesignPatterns
 
Composite Application Library, Prism v2
Composite Application Library, Prism v2Composite Application Library, Prism v2
Composite Application Library, Prism v2
 
Conditional Execution - A Pattern for the Implementation of Fine-Grained Vari...
Conditional Execution - A Pattern for the Implementation of Fine-Grained Vari...Conditional Execution - A Pattern for the Implementation of Fine-Grained Vari...
Conditional Execution - A Pattern for the Implementation of Fine-Grained Vari...
 
COCOMA presentation, FIA 2013
COCOMA presentation, FIA 2013COCOMA presentation, FIA 2013
COCOMA presentation, FIA 2013
 
Software engineering Questions and Answers
Software engineering Questions and AnswersSoftware engineering Questions and Answers
Software engineering Questions and Answers
 
Integrating profiling into mde compilers
Integrating profiling into mde compilersIntegrating profiling into mde compilers
Integrating profiling into mde compilers
 
Configurability for Cloud-Native Applications: Observability and Control
Configurability for Cloud-Native Applications: Observability and ControlConfigurability for Cloud-Native Applications: Observability and Control
Configurability for Cloud-Native Applications: Observability and Control
 
Software Engineering CSE/IT.pptx
 Software Engineering CSE/IT.pptx Software Engineering CSE/IT.pptx
Software Engineering CSE/IT.pptx
 
Object oriented framework
Object oriented frameworkObject oriented framework
Object oriented framework
 
Adapting Applications on the Fly
Adapting Applications on the FlyAdapting Applications on the Fly
Adapting Applications on the Fly
 
Ch16-Software Engineering 9
Ch16-Software Engineering 9Ch16-Software Engineering 9
Ch16-Software Engineering 9
 
Module 3.1.pptx
Module 3.1.pptxModule 3.1.pptx
Module 3.1.pptx
 
term paper for cbd models
term paper for cbd modelsterm paper for cbd models
term paper for cbd models
 
Learning Software Performance Models for Dynamic and Uncertain Environments
Learning Software Performance Models for Dynamic and Uncertain EnvironmentsLearning Software Performance Models for Dynamic and Uncertain Environments
Learning Software Performance Models for Dynamic and Uncertain Environments
 
CS587 Project - Raychaudhury,Shaalmali
CS587 Project - Raychaudhury,ShaalmaliCS587 Project - Raychaudhury,Shaalmali
CS587 Project - Raychaudhury,Shaalmali
 
SSE Integrations Overview
SSE Integrations OverviewSSE Integrations Overview
SSE Integrations Overview
 

Dernier

+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
?#DUbAI#??##{{(☎️+971_581248768%)**%*]'#abortion pills for sale in dubai@
 
Pests of cotton_Sucking_Pests_Dr.UPR.pdf
Pests of cotton_Sucking_Pests_Dr.UPR.pdfPests of cotton_Sucking_Pests_Dr.UPR.pdf
Pests of cotton_Sucking_Pests_Dr.UPR.pdf
PirithiRaju
 
The Mariana Trench remarkable geological features on Earth.pptx
The Mariana Trench remarkable geological features on Earth.pptxThe Mariana Trench remarkable geological features on Earth.pptx
The Mariana Trench remarkable geological features on Earth.pptx
seri bangash
 
Bacterial Identification and Classifications
Bacterial Identification and ClassificationsBacterial Identification and Classifications
Bacterial Identification and Classifications
Areesha Ahmad
 
development of diagnostic enzyme assay to detect leuser virus
development of diagnostic enzyme assay to detect leuser virusdevelopment of diagnostic enzyme assay to detect leuser virus
development of diagnostic enzyme assay to detect leuser virus
NazaninKarimi6
 

Dernier (20)

+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
 
chemical bonding Essentials of Physical Chemistry2.pdf
chemical bonding Essentials of Physical Chemistry2.pdfchemical bonding Essentials of Physical Chemistry2.pdf
chemical bonding Essentials of Physical Chemistry2.pdf
 
Introduction to Viruses
Introduction to VirusesIntroduction to Viruses
Introduction to Viruses
 
Zoology 5th semester notes( Sumit_yadav).pdf
Zoology 5th semester notes( Sumit_yadav).pdfZoology 5th semester notes( Sumit_yadav).pdf
Zoology 5th semester notes( Sumit_yadav).pdf
 
Pests of cotton_Sucking_Pests_Dr.UPR.pdf
Pests of cotton_Sucking_Pests_Dr.UPR.pdfPests of cotton_Sucking_Pests_Dr.UPR.pdf
Pests of cotton_Sucking_Pests_Dr.UPR.pdf
 
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
 
The Mariana Trench remarkable geological features on Earth.pptx
The Mariana Trench remarkable geological features on Earth.pptxThe Mariana Trench remarkable geological features on Earth.pptx
The Mariana Trench remarkable geological features on Earth.pptx
 
Human & Veterinary Respiratory Physilogy_DR.E.Muralinath_Associate Professor....
Human & Veterinary Respiratory Physilogy_DR.E.Muralinath_Associate Professor....Human & Veterinary Respiratory Physilogy_DR.E.Muralinath_Associate Professor....
Human & Veterinary Respiratory Physilogy_DR.E.Muralinath_Associate Professor....
 
300003-World Science Day For Peace And Development.pptx
300003-World Science Day For Peace And Development.pptx300003-World Science Day For Peace And Development.pptx
300003-World Science Day For Peace And Development.pptx
 
GBSN - Microbiology (Unit 2)
GBSN - Microbiology (Unit 2)GBSN - Microbiology (Unit 2)
GBSN - Microbiology (Unit 2)
 
pumpkin fruit fly, water melon fruit fly, cucumber fruit fly
pumpkin fruit fly, water melon fruit fly, cucumber fruit flypumpkin fruit fly, water melon fruit fly, cucumber fruit fly
pumpkin fruit fly, water melon fruit fly, cucumber fruit fly
 
Grade 7 - Lesson 1 - Microscope and Its Functions
Grade 7 - Lesson 1 - Microscope and Its FunctionsGrade 7 - Lesson 1 - Microscope and Its Functions
Grade 7 - Lesson 1 - Microscope and Its Functions
 
Bacterial Identification and Classifications
Bacterial Identification and ClassificationsBacterial Identification and Classifications
Bacterial Identification and Classifications
 
GBSN - Microbiology (Unit 3)
GBSN - Microbiology (Unit 3)GBSN - Microbiology (Unit 3)
GBSN - Microbiology (Unit 3)
 
Molecular markers- RFLP, RAPD, AFLP, SNP etc.
Molecular markers- RFLP, RAPD, AFLP, SNP etc.Molecular markers- RFLP, RAPD, AFLP, SNP etc.
Molecular markers- RFLP, RAPD, AFLP, SNP etc.
 
FAIRSpectra - Enabling the FAIRification of Analytical Science
FAIRSpectra - Enabling the FAIRification of Analytical ScienceFAIRSpectra - Enabling the FAIRification of Analytical Science
FAIRSpectra - Enabling the FAIRification of Analytical Science
 
development of diagnostic enzyme assay to detect leuser virus
development of diagnostic enzyme assay to detect leuser virusdevelopment of diagnostic enzyme assay to detect leuser virus
development of diagnostic enzyme assay to detect leuser virus
 
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verifiedConnaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
 
Thyroid Physiology_Dr.E. Muralinath_ Associate Professor
Thyroid Physiology_Dr.E. Muralinath_ Associate ProfessorThyroid Physiology_Dr.E. Muralinath_ Associate Professor
Thyroid Physiology_Dr.E. Muralinath_ Associate Professor
 
Proteomics: types, protein profiling steps etc.
Proteomics: types, protein profiling steps etc.Proteomics: types, protein profiling steps etc.
Proteomics: types, protein profiling steps etc.
 

Coordinated run-time adaptation of variability-intensive systems: an application in cloud computing (VACE 2016)

  • 1. ©paluno Coordinated Run-time Adaptation of Variability-intensive Systems An Application in Cloud Computing Andreas Metzger†, Andreas Bayer†, Daniel Doyle*, Amir Molzam Sharifloo†, Klaus Pohl†, Florian Wessling† † paluno (The Ruhr Institute for Software Technology), University of Duisburg Essen, Germany * Intel, Ireland
  • 3. ©paluno Motivation Adaptive Software Systems  Adaptive software systems can modify own structure and behavior at run time to cope with dynamic changes in…  M = machine (software)  self-healing  W = world (context)  context-aware  R = changing requirements  ???  @runtime: adaptive systems monitor changes in M and W or even R directly  M, W |≠ R (requirements violation)  self-modification 3 M, W |= R ?
  • 4. ©paluno Motivation Coordination among Adaptive Systems  Distributed systems (e.g., cloud systems or cyber-physical systems) orchestrate many adaptive sub-systems  Each sub-system may perform adaptations simultaneously and independent of each other  However, adaptations may affect shared phenomena, thus:  Conflicts between adaptations may occur  Synergies among adaptations may be missed 4 Shared Phenomenon M, W |= R ? M, W |= R ? M, W |= R ?
  • 5. ©paluno Motivation Use Case: Conflicts in Adaptive Cloud Systems  „AdvancedTV“ Use Case from EU Project Cloud Wave  Cloud application that offers services in parallel to running TV programme  Two adaptive systems:  Cloud Infrastructure (IaaS: CPU, RAM, …)  Cloud Application (SaaS)  Adaptations of Cloud Infrastructure  Horizontal Scaling  E.g., turning off virtual machines to save energy  Vertical Scaling  …  Adaptations of Cloud Application  Different levels of social media features  No  Partial  Unlimited  … 5 Performance -- VM  -- Performance ++ Socia Media  -- Performance
  • 7. ©paluno FCORE Approach Main Ideas and Challenges  Explicitly model adaptations and dependencies among systems during design time  Challenge 1: Developers must model adaptations of their systems  Sufficiently compact, yet expressive modeling technique  Challenge 2: Systems developed by different developers/organizations  Suitably (small) common denominator to describe dependencies among systems  Analyze models at run time to determine conflicts and identify optimizations (synergies)  Challenge 3: Self-adaptation at run time must be fast enough to be effective (otherwise may be too late)  Efficient model analysis during system execution 7
  • 9. ©paluno FCORE Models Main Underlying Concepts  FCORE = DSPL Feature Model + Goal Models 9 Concepts from Dynamic Software Product Lines Feature Models to describe adaptations Concepts from Goal Models to describe dependencies via shared phenomena Main underlying assumption: “Known Unkowns!“
  • 10. ©paluno FCORE Models Why DSPL Models? (Challenge 1)  DSPLs can build on proven engineering foundations of SPLs!  DSPL extend existing software product line engineering approaches by moving their capabilities to run time  Variability binding is postponed to run time, allowing a DSPL to activate or deactivate certain features  Configurations of a DSPL are expressed in terms of a product line variability model, usually a feature model 10 Classical SPL Dynamic SPL variability describes different pos- sible software systems variability describes different possible configurations (i.e., adaptations) of the same system
  • 11. ©paluno FCORE Models Which Kinds of DSPL Models? (Challenges 1&3) Approach Expressiveness Analysis Basic-FM High redundancy in models (replication of FM sub-trees) Cardinalities only 1..1 / 0..n SAT solver Cardinality-Based FM Alternative-Groups Cardinalities n..m Feature-Cardinalities n..m (i.e., instantiation of features) SAT solver Extended-FM Feature-Attributes (Integer, Enumeration, …) CSP solver 11
  • 12. ©paluno FCORE Models Why Goal Models? (Challenge 2)  Soft Goals provide high-level of abstraction to describe influences of features (~ “tasks”) on goal satisfaction  Well-known from requirements engineering  Defining dependencies among systems requires agreeing on a set of shared soft goals  E.g., in cloud computing, these soft goals may be derived from standardized QoS catalogues for SLAs 12
  • 13. ©paluno FCORE Models CloudWave Use Case: Simplified 13
  • 17. ©paluno FCORE Analysis Main Underlying Strategy (Challenge 3)  Formalize FCORE Model as CSP (justification see above)  Perform automated reasoning on formalization  Two main usages:  FCORE Filter: Validity check of given configurations (= detecting conflicts)  E.g., 1 CPU + Unlimited Social Media  violation of high performance  FCORE Search: Search for configurations with high goal satisfaction (= exploiting synergies)  E.g., 6 CPUs + No Social Media  high performance + low costs 17
  • 18. ©paluno FCORE Analysis Formalization: Features 18 A = Feature selected  Feature  Requires-Relation  Excludes-Relation  Feature Group
  • 19. ©paluno FCORE Analysis Formalization: Goals 20  Softgoals and Attributes Softgoal satisfaction: sgVal = [-1.0, +1.0]
  • 20. ©paluno FCORE Analysis Performance (Challenge 3)  FCORE Filter  No performance issues  Just compute goal satisfaction for given configuration  Ca. 2ms for cloud use case  FCORE Search  CSP to find optimal configurations (maximize sgVal)  Experimental results for cloud use case 21
  • 22. ©paluno Conclusion and Outlook  Concluded: FCORE as an approach for coordinating among adaptive, variability intensive systems  Building on DSPLs  Offering Modelling + Analysis  Exemplified for the case of cloud computing  Ongoing:  Implementation as part of CloudWave Adaptation Engine (jointly with IBM and intel)  Future:  Handling “Unknown Unknowns”: Extending DSPLs with dynamic learning and evolution 23
  • 23. ©paluno The research leading to these results has received funding from the European Union's Seventh Framework Programme FP7/2007- 2013 under grant agreement 610802 (CloudWave) http://www.cloudwave-fp7.eu/ Thank You!