Autonomic Performance-Aware Resource Management in Dynamic IT Service Infrastructures

Autonomic Performance-Aware Resource Management in Dynamic IT Service Infrastructures

escartes Autonomic Performance-Aware Resource Management in Dynamic IT Service Infrastructures Zur Erlangung des akademischen Grades eines Doktors der Ingenieurwissenschaften/ Doktors der Naturwissenschaften der Fakultät für Informatik des Karlsruher Instituts für Technologie (KIT) genehmigte Dissertation von Nikolaus Matthias Huber aus Zell am Harmersbach Tag der mündlichen Prüfung: 16.07.2014 Erster Gutachter: Prof. Dr. Samuel Kounev Zweiter Gutachter: Prof. Dr. Ralf Reussner KIT – Universität des Landes Baden-Württemberg und nationales Forschungszentrum in der Helmholtz-Gemeinschaft www.kit.edu Ich versichere wahrheitsgem¨aß, die Dissertation bis auf die dort angegebenen Hilfen selb- st¨andig angefertigt, alle benutzten Hilfsmittel vollst¨andig und genau angegeben und alles kenntlich gemacht zu haben, was aus Arbeiten anderer und eigenen Ver¨offentlichungen unver¨andert oder mit Anderungen¨ entnommen wurde. Karlsruhe, 30. September 2014 Nikolaus Huber Contents Abstract xiii Zusammenfassung xv 1. Introduction 1 1.1. Motivation . .1 1.2. Problem Statement . .2 1.3. Shortcomings of Existing Approaches . .4 1.4. Contributions of this Thesis . .6 1.5. Outline . .8 I. Foundations and Related Work 11 2. Autonomic System Adaptation and Performance Modeling 13 2.1. Autonomic and Self-Adaptive Systems . 14 2.1.1. Self-* Properties . 14 2.1.2. Design Principles of Autonomic Systems . 16 2.1.3. Model-Driven Autonomic and Self-Adaptive Systems . 18 2.1.4. The Descartes Research Project and Self-Aware Computing Systems 20 2.2. Software Performance Engineering . 21 2.2.1. Performance Models: Classification and Terminology . 21 2.2.2. Model-Driven Software Performance Engineering . 22 2.2.3. The Palladio Component Model (PCM) . 22 2.2.4. Online Performance Prediction . 24 3. Related Work 29 3.1. Architecture-Based Self-Adaptive Software Systems . 30 3.1.1. Engineering and Evaluation of Self-Adaptive Software Systems . 30 3.1.2. Architecture-Based Self-Adaptation Approaches . 32 3.2. Model-Based Performance and Resource Management . 36 3.2.1. Predictive Performance Models . 36 3.2.2. Architecture-Level Performance Models . 37 3.3. Adaptation Process Modeling Languages . 37 3.4. Summary . 38 II. Proactive Model-Based Performance and Resource Management 41 4. Proactive Model-Based System Adaptation 43 4.1. Model-Based Adaptation Control Loop . 44 4.1.1. Monitor . 44 v vi Contents 4.1.2. Analyze . 45 4.1.3. Plan . 46 4.1.4. Execute . 47 4.1.5. Knowledge . 47 4.2. Overview of the Descartes Modeling Language . 47 4.2.1. Technical Viewpoint . 49 4.2.2. Logical Viewpoint . 51 4.3. Summary . 52 5. Modeling System Resource Landscapes and Their Performance Influences 53 5.1. Resource Landscape Meta-Model . 53 5.1.1. Containers and Containment Relationships . 54 5.1.2. Classes of Runtime Environments . 55 5.1.3. Resource Configuration Specification . 56 5.1.4. Container Types . 57 5.1.5. Example Resource Landscape Model Instance . 59 5.2. Performance-Influencing Factors of Resource Layers . 60 5.2.1. Classification of Performance-Influencing Factors . 60 5.2.2. Automatic Quantification of Performance-Influencing Factors . 62 5.2.3. Derivation of the Performance Model . 75 5.3. Application Architecture, Usage Profile, and Deployment Meta-Models . 76 5.3.1. Application Architecture Meta-Model . 76 5.3.2. Usage Profile Meta-Model . 77 5.3.3. Deployment Meta-Model . 78 5.4. Case Studies . 78 5.4.1. Modeling Data Centers with the Resource Landscape Meta-Model . 78 5.4.2. Quantifying Performance Influences of Virtualization . 81 5.5. Summary . 84 6. Modeling System Adaptation Processes 85 6.1. Adaptation Points Meta-Model . 86 6.2. Adaptation Process Meta-Model . 89 6.2.1. Actions . 91 6.2.2. Tactics . 93 6.2.3. Strategies . 95 6.2.4. QoS Data Repository . 96 6.2.5. Weighting Function . 97 6.3. Adaptation Framework Architecture and Implementation . 99 6.4. Evaluation . 103 6.4.1. Comparing S/T/A, Story Diagrams, and Stitch . 103 6.4.2. Comparing Accuracy and Efficiency Using PerOpteryx . 104 6.4.3. Reusing Adaptation Plans in SLAstic . 107 6.5. Summary . 109 7. Self-Adaptive Workload Classification and Forecasting 111 7.1. Time Series Analysis . 111 7.1.1. Workload Intensity Behavior Characteristics . 112 7.1.2. Survey of Forecasting Methods . 114 7.2. Workload Classification and Forecasting . 118 7.2.1. Forecasting Objectives . 119 7.2.2. Forecasting Methods Overhead Groups . 120 7.2.3. Forecasting Methods Partitions . 120 vi Contents vii 7.2.4. Evaluating Forecasting Accuracy . 121 7.2.5. Non-Absolutely Positive Workloads . 122 7.2.6. Decision Tree . 122 7.3. WCF Architecture and Implementation . 123 7.4. Evaluation . 126 7.4.1. Experiment Design . 127 7.4.2. Experiment I: Comparing WCF with ETS and Naive Forecasting . 127 7.4.3. Experiment II: Comparing WCF with Overhead Group 2 . 130 7.4.4. Experiment III: Comparing WCF with Overhead Group 3 . 134 7.4.5. Experiment IV: Comparing WCF with Overhead Group 4 . 138 7.5. Summary . 142 III. Validation and Conclusion 145 8. Validation 147 8.1. Validation Goals . 148 8.1.1. Modeling Capabilities . 148 8.1.2. Prediction Capabilities of the Architecture-Level Performance Model 148 8.1.3. End-to-End Validation of the Model-Based Adaptation Approach . 149 8.2. Model-Based Resource Allocation in Virtualized Environments . 151 8.2.1. SPECjEnterprise2010 Benchmark and Adaptation Process Overview 151 8.2.2. Applied DML Instance . 156 8.2.3. Evaluation . 160 8.3. Proactive Model-Based System Adaptation . 165 8.3.1. Proactively Reducing SLA Violations with WCF . 166 8.3.2. WCF for Proactive Resource Provisioning at Run-Time . 170 8.3.3. Evaluation . 172 8.4. Model-Based System Adaptation in Heterogeneous Environments . 172 8.4.1. Blue Yonder System Architecture . 173 8.4.2. Applied DML Instance . 175 8.4.3. Evaluation . 180 8.5. Discussion . 182 9. Conclusions and Outlook 187 9.1. Summary . 187 9.2. Outlook . 189 Appendix A. Additional Meta-Model Specifications 193 A.1. Application Architecture Meta-Model . 193 A.1.1. Service Behavior Descriptions . 193 A.1.2. Signature . 195 A.2. Usage Profile Meta-Model . 195 List of Figures 199 List of Tables 203 Bibliography 205 vii Publication List [Huber et al., 2014] Huber, N., van Hoorn, A., Koziolek, A., Brosig, F., and Kounev, S. (2014). Modeling Run-Time Adaptation at the System Architecture Level in Dynamic Service-Oriented Environments. Service Oriented Computing and Applications Journal (SOCA), 8(1):73–89. [Huber et al., 2012d] Huber, N., von Quast, M., Brosig, F., Hauck, M., and Kounev, S. (2012d). A Method for Experimental Analysis and Modeling of Virtualization Perfor- mance Overhead. In Ivanov, I., van Sinderen, M., and Shishkov, B., editors, Cloud Computing and Services Science, Service Science: Research and Innovations in the Ser- vice Economy, pages 353–370. Springer, New York. [Huber et al., 2012c] Huber, N., van Hoorn, A., Koziolek, A., Brosig, F., and Kounev, S. (2012c). S/T/A: Meta-Modeling Run-Time Adaptation in Component-Based System Architectures. In Proceedings of the 9th IEEE International Conference on e-Business Engineering (ICEBE 2012), pages 70–77, Los Alamitos, CA, USA. IEEE Computer Society. Acceptance Rate (Full Paper): 19.7% (26/132). [Huber et al., 2012b] Huber, N., Brosig, F., and Kounev, S. (2012b). Modeling Dynamic Virtualized Resource Landscapes. In Proceedings of the 8th ACM SIGSOFT Interna- tional Conference on the Quality of Software Architectures (QoSA 2012), pages 81–90, New York, NY, USA. ACM. Acceptance Rate (Full Paper): 25.6%. [Huber et al., 2012a] Huber, N., Brosig, F., Dingle, N., Joshi, K., and Kounev, S. (2012a). Providing Dependability and Performance in the Cloud: Case Studies. In Wolter, K., Avritzer, A., Vieira, M., and van Moorsel, A., editors, Resilience Assessment and Eval- uation of Computing Systems, XVIII. Springer-Verlag, Berlin, Heidelberg. ISBN: 978- 3-642-29031-2. [Huber et al., 2011b] Huber, N., von Quast, M., Hauck, M., and Kounev, S. (2011b). Eval- uating and Modeling Virtualization Performance Overhead for Cloud Environments. In Proceedings of the International Conference on Cloud Computing and Services Science (CLOSER 2011), pages 563 – 573. SciTePress. Acceptance Rate: 18/164 = 10.9%, Best Paper Award. [Huber et al., 2011a] Huber, N., Brosig, F., and Kounev, S. (2011a). Model-based Self- Adaptive Resource Allocation in Virtualized Environments. In Proceedings of the 6th International Symposium on Software Engineering for Adaptive and Self-Managing Sys- tems (SEAMS 2011), pages 90–99, New York, NY, USA. ACM. Acceptance Rate (Full Paper): 27% (21/76). [Huber et al., 2010b] Huber, N., von Quast, M., Brosig, F., and Kounev, S. (2010b). Anal- ysis of the Performance-Influencing Factors of Virtualization Platforms. In Proceedings of the 12th International Symposium on Distributed Objects, Middleware, and Appli- cations (DOA 2010), Crete, Greece. Springer Verlag. Acceptance Rate (Full Paper): 33%. ix x Publication List [Huber et al., 2010a] Huber, N., Becker, S., Rathfelder, C., Schweflinghaus, J., and Reuss- ner, R. (2010a). Performance Modeling in Industry: A Case Study on Storage Virtual- ization. In Proceedings of the ACM/IEEE 32nd International Conference on Software Engineering (ICSE 2010), Software Engineering in Practice Track, pages 1–10, New York, NY, USA. ACM. Acceptance Rate (Full Paper): 23% (16/71). [Huber, 2009] Huber, N. (2009). Performance Modeling of Storage Virtualization. Master’s thesis, Universit¨at Karlsruhe (TH), Karlsruhe, Germany. GFFT Prize. [Brosig et al., 2013a] Brosig, F., Gorsler, F., Huber, N., and Kounev, S. (2013a). Evaluat- ing Approaches for Performance Prediction in Virtualized Environments. In Proceedings of the IEEE 21st International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems (MASCOTS 2013). (Short Paper). [Brosig et al., 2011] Brosig, F., Huber, N., and Kounev, S. (2011). Automated Extraction of Architecture-Level Performance Models of Distributed Component-Based Systems.

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    233 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us