Descartes - Introduction and Status Update Prof. Samuel Kounev, Uni - PowerPoint PPT Presentation

Descartes - Introduction and Status Update Prof. Samuel Kounev, Uni Würzburg Chair of Software Engineering University of Würzburg http://se.informatik.uni-wuerzburg.de/ http://descartes.tools 1 pmw.fortiss.org Munich, November 5th, 2015

Self-Aware Resource Management Response time Response time Service Level Service Level Agreement Agreement Online prediction of SLA violation t 0 t 0 Time Time Online prediction of reconfiguration impact 2 pmw.fortiss.org Munich, November 5th, 2015 2

Descartes Tool Chain http://descartes.tools 3 pmw.fortiss.org Munich, November 5th, 2015 3

Selected Tools  DML – Descartes Modeling Language (homepage, publications)  DML Bench (homepage, publications) DQL – Declarative performance query language (homepage, publications)   LibReDE - Library for resource demand estimation (homepage, publications) LIMBO – Load intensity modeling tool (homepage, publications)  WCF – Workload classification & forecasting tool (homepage, publications)  BUNGEE – Elasticity benchmarking framework (homepage, publications)  hInjector – Security benchmarking tool (homepage, publications)   Queueing Petri Net Modeling Environment (QPME)  Further relevant research  http://descartes-research.net/research/research_areas/ 4 pmw.fortiss.org Munich, November 5th, 2015 4  Self Aware Computing (publications)

Descartes Modeling Language (DML)  Problem: How to model the performance and resource management related aspects of an IT system to enable self-aware resource management? http://descartes.tools/dml 5 pmw.fortiss.org Munich, November 5th, 2015 5

DNI – Network Infrastructure Modeling  Language for modeling of data center networks including SDN-based infrastructures  network topology, switches, routers, virtual machines, network protocols, routes, flow-based configuration,...  Model solvers based on simulation (OMNeT) http://descartes.tools/dni 6 pmw.fortiss.org Munich, November 5th, 2015 6

LIMBO Tool  Problem:  How to capture the load intensity variations (e.g., requests per sec) in a compact mathematical model?  How to forecast the load intensity (requests per sec) in future time horizons?  Load Intensity Modeling & Forecasting Tool http://descartes.tools/limbo 8 pmw.fortiss.org Munich, November 5th, 2015 8

LIMBO Tool  Workload Classification & Forecasting (WCF)  Use of multiple alternative forecasting methods in parallel  Selection of method based on its accuracy in the past workload intensity history now near future http://descartes.tools/wcf 9 pmw.fortiss.org Munich, November 5th, 2015 9

LibReDE Tool  Problem: How to estimate the total service time of a given type of request/job at a given resource?  Library for Resource Demand Estimation  Ready-to-use implementations of estimation approaches  Selection of a suitable approach for a given scenario http://descartes.tools/librede S. Spinner, G. Casale, F. Brosig, and S. Kounev. Evaluating Approaches to Resource Demand Estimation . Performance Evaluation , 92:51 - 71, October 2015, Elsevier B.V. [ DOI | http | .pdf ] 10 pmw.fortiss.org Munich, November 5th, 2015 10

Systems Benchmarking Metrics and benchmarks for quantitative evaluation of 1. Cloud elasticity 2. Performance isolation 3. Security (Intrusion detection and prevention) 4. ... http://research.spec.org S. Kounev. Quantitative Evaluation of Service Dependability in Shared Execution Environments (Keynote Talk). In 11th Intl. Conf. on Quantitative Evaluation of SysTems (QEST 2014), Florence, Italy, September 8-12, 2014. [ slides | extended abstract ] [geek & poke] 11 pmw.fortiss.org Munich, November 5th, 2015 11

BUNGEE Tool  Problem: How to measure and quantify cloud elasticity?  Framework for benchmarking elasticity  Current focus: IaaS cloud platforms http://descartes.tools/bungee 12 pmw.fortiss.org Munich, November 5th, 2015 12

hInjector Tool MVM! Hypervisor! Con fi guration! Logs !  Cooperation with User! 1! 6! Injector! Kernel! Filter!  Netflix, Inc. LKM! 2! ! Hypercall 3! 4! handler! 5!  Siemens Corporate Research Memory! Hardware! vCPU! shared_info! monitors! 3! IDS ! 5! (in SVM)!  ERNW  Benchmarking security of virtualized infrastructures  Vulnerability analysis  Evaluation of security mechanisms (intrusion detection techniques) Academic Partner 13 pmw.fortiss.org Munich, November 5th, 2015 13

Self-Aware Resource Management Response time Response time Service Level Service Level Agreement Agreement Online prediction of SLA violation t 0 t 0 Time Time Online prediction of reconfiguration impact Example of Self-Aware Computing See http://www.dagstuhl.de/15041 Dagstuhl Seminar 15041, January 18-23, 2015 14 pmw.fortiss.org Munich, November 5th, 2015 14

Definition Self-aware Computing Systems are computing systems that: 1. learn models capturing knowledge about themselves and their environment on an ongoing basis and 2. reason using the models enabling them to act based on their knowledge and reasoning in accordance with higher-level goals , which may also be subject to change. S. Kounev, X. Zhu, J. O. Kephart and M. Kwiatkowska, editors. Model-driven Algorithms and Architectures for Self-Aware Computing Systems (Dagstuhl Seminar 15041). Dagstuhl Reports, vol. 5, No. 1. pp. 164-196, Dagstuhl, Germany, 2015. http://drops.dagstuhl.de/opus/volltexte/2015/5038 Community page: http://descartes.tools/self-aware 15 pmw.fortiss.org Munich, November 5th, 2015 15

Self-Aware Learning & Reasoning Loop 16 pmw.fortiss.org Munich, November 5th, 2015 16

Descartes Talks  Thu 13:00: Jóakim von Kistowski. Common Errors and Assumptions in Energy Measurement and Management  Thu 14:00: Simon Spinner. Resource demand estimation in distributed, service-oriented applications using LibReDE  Thu 17:00: Andreas Weber. BUNGEE: An Elasticity Benchmark for Self-Adaptive IaaS Cloud Environments  Fri 13:20: Jürgen Walter und Simon Eismann Automated Transformation of DML to PCM models 17 pmw.fortiss.org Munich, November 5th, 2015 17