120 likes | 239 Views
Distributed Autonomic Management (DAM). Nitin Bande . Introduction DAM concept DAM Approach Related Work Summary Conclusion. Introduction. Current approaches to network model employ client/server model, non-intelligent mobile agents.
E N D
Distributed Autonomic Management (DAM) Nitin Bande
Introduction • DAM concept • DAM Approach • Related Work • Summary • Conclusion
Introduction • Current approaches to network model employ client/server model, non-intelligent mobile agents. • Autonomic Computing: Initiative was proposed by IBM in 2001 • Inspired by biological systems such as the autonomic human nervous system • Deal with complexity, dynamism, heterogeneity and uncertainty • significant new strategic and holistic approach to the design of complex distributed computer systems • Provides the user with an interface that exactly meets her/his needs
DAM Concepts • Distributed Autonomic Computing & Autonomic Nervous System • Human Nervous system controls the vegetative functions of the body • Many decisions made by autonomic elements in body are involuntary • Biological self-management is influenced for developing self-management within the systems • Distinction between autonomic activity in the human body and autonomic responses in computer systems • to cope with the rapidly growing complexity of integrating, managing, and operating computing system • Autonomic elements in computer systems make decisions based on tasks which are chosen to be delegated to the technology
Distributed Autonomic Computing • Characteristics • Self-Configuration : Ability of the system to automatically adapt to changes • Self-Healing : Ability of the system to discover, diagnose & react to disruptions • Self-Optimization : Ability of the system to maximize the resource utilization • Self-Protection : Ability of the system to detect, diagnose & act to prevent disruptions • Self-Awareness : System is aware of its states and behavior • Benefits • Efficiency, Maintainability, Functionality, Reliability, Usability, and Portability
DAM Reference Model • Managed elements • Sensors • Autonomic Manager • Effectors
AI & DAM Components • Soft computing techniques • Neural networks, fuzzy logic, probabilistic reasoning incorporating and so on • Machine learning techniques, optimization techniques, fault diagnosis techniques, feedback control, and planning techniques • Clockwork • A method provides predictive self-management, regulates behavior in anticipation of need using statistical modeling, tracking and forecasting methods, Self-configuration element • Probabilistic technique • Autonomic algorithm selection • Self-training and self-optimization to find the best algorithm
Large-scale server management and control • Time-series methods, rule-based classification for a self- management and control system • Calculation of costs in an autonomic system and the self- healing equation • Machine design • Reaction • The lowest level where no learning occurs and only involves immediate response to state information coming from sensory systems • Routine • Middleware level where largely routine evaluation and planning behaviors take place • Reflection • Top level, which receives input from below • Meta process, where the mind deliberates about itself
DAM Approaches • Stationary intelligent agent approach & Mobile agent approach • Endowing traditional SNMP agents that were essential in the client/server model with some form of intelligence. • Collaboration of SNMP agents with mobile agents. • Bandwidth problem is the primary focus. • Detection of faults and performance degradations of distributed networks. • Agents may share knowledge by sharing cases or by sharing adaptation procedures.
Challenges of DAM • Identification and accessibility • Analyze monitored data • Self-configuration in large scale applications • Problem localization • Decision making • Self protecting against active threats
Conclusion • Inspired by biological systems such as the autonomic human nervous system • Enables the development of self-managing computing systems and applications • The systems/applications use autonomic strategies and algorithms to handle complexity and uncertainties with minimum human intervention • Implement intelligent control loops to monitor, analyze, plan and execute using knowledge of the environment • Challenge is be accomplished through a combination of process changes, skills evolution, new technologies and architecture, and open industry standards