Survey of Validation Methods in Autonomic Computing Systems Ronald Stevens and Brittany Parsons June 15, 2006 REU sponsored by NSF Outline of Presentation 1. What is an autonomic computing system? 2. History of autonomic computing 3. Characteristics of ACSs 4. Architecture of ACSs 5. Overview and Evaluation of projects 6. Future Work 7. References 2 1. What is an autonomic computing system (ACS)? Autonomic computing is a computing environment with the ability to manage itself and dynamically adapt to change in accordance with business policies and objectives [IBM05] 3 2. History of ACSs • IBM first introduced the idea of Autonomic Computing in March 2001. • The word autonomic was used to describe this computing because of its biological origins. • Autonomic systems of the body handle involuntary functions that are needed for the body to live. 4 3. Characteristics of ACSs - Diagnose and correct problems - Automatically adapts to change Self Configuring Self Healing Self Optimizing Self Protecting - Able to - Be able to track anticipate and changes and act handle security accordingly 5 Figure 1 Diagram of Autonomic Computing risks Characteristics [IBM05] 4. Architecture of an ACS Element Analyze Monitor Plan Knowledge Sensors monitoring - collects detailed data. Effectors Execute analysis - uses the data gathered. planning – comes up with a plan of action executing - perform desired actions Managed Element Figure 2 Autonomic Manager [KC03] 6 5. Overview and Evaluation of projects • • • • Impala OceanStore Model-Driven Autonomic Manager Bison 7 Evaluation of projects • We evaluated each of the four projects based on the criteria and a scale that we came up with for Autonomic Computing Systems. 8 Evaluation criteria for ACSs • QoS – How well systems perform required tasks • Cost – cost effectiveness • Adaptivity – ability to adapt • Fault Tolerance – ability to deal with problems • Fault Avoidance – ability to avoid problems • Fault Detection – ability to find problems • Degree of Autonomy – amount of human input needed • Granularity – flexibility of system • Availability of Source Code 9 Scale for Criteria Evaluation Rating Description 1 This system does not fit the criteria. 2 This system may possibly fit some part of the criteria. 3 This system has characteristics of the criteria. 4 This system is close to fitting the criteria. 5 This system fits the criteria. 10 Impala • Project Name: Impala • Organization: Princeton University • Research Team: Margaret Martonosi (Lead), Steve Lyon, Li-Shiuan Peh, Vince Poor, Dan Rubenstein, Chris Sadler, Philo Juang, Ting Liu, Yong Wang, Pei Zhang • Purpose of Project: ZebraNet will create a wireless ad-hoc network capable of tracking the movements of wild zebras via tracking collars. [MLP+06] Impala is the middleware that allows each zebra’s collar to act as a node in a wireless network. 11 Impala Adaptivity Cost Quality of Service Granularity Fault Avoidance Fault Tolerance Fault Detection Degree of Autonomy Availability of Source Code Self-Healing 1 1 1 1 1 1 1 1 Self-Protecting 1 1 1 1 1 1 1 1 Self-Configuring Self Optomizing 4 4 1 1 3 3 3 3 4 2 4 2 4 1 5 4 no 12 OceanStore • Project Name: OceanStore • Organization: UC Berekley • Research Team: John Kubiatowicz(Lead), Byung-Gon Chun,Steven Czerwinski, Patrick Eaton • Purpose of Project: OceanStore is designed to be a massively scalable storage system. OceanStore employs promiscuous caching and cryptography to ensure that data will not be lost or compromised. [KGC06] 13 OceanStore Adaptivity Cost Quality of Service Granularity Fault Avoidance Fault Tolerance Fault Detection Degree of Autonomy Availability of Source Code Self-Healing 1 1 1 1 1 1 1 1 Self-Protecting 4 1 4 3 5 5 3 4 Self-Configuring Self Optomizing 3 1 1 1 5 1 3 1 5 1 5 1 2 1 4 1 yes 14 Model-Driven Autonomic Manager • Project Name: Model Driven Autonomic Manager • Organization: Indiana University • Research Team: Dr. Yuashun Dai (Lead) • Purpose of Project: The purpose of this project is to develop a model driven autonomic manager. This project focuses on autonomic traits in networks.[DAI06] 15 Model Driven Autonomic Manager Adaptivity Cost Quality of Service Granularity Fault Avoidance Fault Tolerance Fault Detection Degree of Autonomy Availability of Source Code Self-Healing 5 1 1 5 2 4 4 5 Self-Protecting 5 1 1 5 2 4 4 4 Self-Configuring Self Optomizing 5 5 1 1 1 1 4 5 2 2 4 4 4 4 5 5 no 16 Bison • Project Name: The Bison Project • Organization: Future & Emerging Technologies initiative of the Information Society Technologies Programme of the European Commission. • Research Team: The leads for their resepective universities are Ozalp Babaoglu, Geoff Canwright, Luca Maria Gambardella, and Andreas Deutsch. • Purpose of Project: The Bison Project stands for Biology-Inspired techniques for Self-Organization in dynamic Networks. This project studying biological systems to improve Peer to Peer networks. [DDG05] 17 The Bison Project Adaptivity Cost Quality of Service Granularity Fault Avoidance Fault Tolerance Fault Detection Degree of Autonomy Availability of Source Code Self-Healing 5 1 3 5 5 5 5 5 Self-Protecting 5 1 3 5 5 5 5 5 Self-Configuring Self Optomizing 5 5 1 1 3 3 5 5 4 5 4 5 5 4 4 5 yes 18 6. Future work • Validating autonomic characteristics in a system. 1. 2. 3. 4. 5. 6. 7. Project in conjunction with graduate students in the Software Testing Research Group. Identify a system (X) with a test history (test cases and actual output) Use the IBM Autonomic Toolkit to log the state of X during execution (Overview of the IBM Autonomic Toolkit) Introduce testing concept. Investigate if the logging facility provided by IBM Autonomic Toolkit (IAT) allows for the validation of the X using the test history. Simulate a change for an autonomic characteristic e.g., self configuration, then use the logs generated by the IAT and the test history of X to check if the behavior of X can be validated. To accomplish the above we will be learning how to use two automated testing tools – JUnit and Rationla Functional Tester. 19 7. References [IBM05] IBM Corporation. IBM, 2005. An architectural blueprint for autonomic computing. June 2005. [KC03] J. Kephart and D. Chess. The vision of autonomic computing. Computer, vol. 36, no.1, pp. 41-50, January 2003. [MH04] J. McCann and M. Huebscher. Evaluation issues in autonomic computing. International Workshop on Agents and Autonomic Computing and Grid Enabled Virtual Organizations (AACGEVO04), Wuhan, China 2004 [MLP+06] ZebraNet. http://www.princeton.edu/~mrm/zebranet.html [KGC06] Ocean Store. http://oceanstore.cs.berkeley.edu/info/overview.html [DAI06] Y. Dai. Model-Driven Autonomic Management. 2006 http://www.cs.iupui.edu/~ydai/AMSCS/MAM.htm [DDG05] G. Di Caro, F. Ducatelle, and L.M. Gambardella. Project description: BISON: Biology-Inspired techniques for SelfOrganization in dynamic Networks. Zeitschrift Knstliche Intelligenz, Special Issue on Swarm Intelligence, November 2005. 20