Transcript Autonomic Computing

CS600 Distributed Systems, Assignment #5
Autonomic Computing
Nov. 22. 2006
Eric Joonmyung Kang
Distributed Processing & Network Management Lab.
Dept. of Computer Science and Engineering
POSTECH, Korea
[email protected]
Autonomic Computing
-1-
DP&NM Lab.
CSE, POSTECH
Presentation Outline
 Introduction
 Autonomic Computing
– Architecture concepts
– Architecture details
 Application
 Research Issues and Challenges
 Conclusion
Autonomic Computing
-2-
DP&NM Lab.
CSE, POSTECH
Growing Complexity…
 Future network environment and system
Billing
ISP
VHE SIP Proxy
Server
Signalling WAP Accounting
Gateway
The
Internet
Context-aware information
Centre
Satellite
Core Network
Broadcast
Networks
(DAB, DVB-T)
access operator
CDMA, GSM,
GPRS
4G
Autonomic
Computing
IP-based
micro-mobility
ASP
Autonomic Computing
DMB
MP3
Wireless
M-banking
LANs
corporation
ISP
WiBro,
HSDPA
Camera
Bluetooth
Zigbee
Navigation
-3-
Cam
DP&NM Lab.
CSE, POSTECH
Growing Complexity…
 Very large scales
– million of entities
 Ad hoc (amorphous) structures/behaviors
– p2p/hierarchical architecture
 Dynamic
– entities join, leave, move, change behavior
 Heterogeneous
– capability, connectivity, reliability, guarantees, QoS
 Unreliable
– components, communication
 Lack of common/complete knowledge
– number, type, location, availability, connectivity, protocols,
semantics, etc.
Autonomic Computing
-4-
DP&NM Lab.
CSE, POSTECH
Convergence of IT and BT
 Our system programming paradigms, methods and
management tools seem to be inadequate for handling
the scale, complexity, dynamism and heterogeneity of
emerging future network and systems
 Biological systems have evolved strategies to cope with
dynamic, complex, highly uncertain constraints
 Recently, some research area of IT is applying biological
system concept to solve its unsolved problems
Autonomic Computing
-5-
DP&NM Lab.
CSE, POSTECH
Autonomic Computing
 Autonomic
– Pertaining to an on demand operating environment that responds
automatically to problems, security threats, and system failures
– Cf) self-adaptive, self-governing, self-managing, …
 The term “autonomic” is derived from human biology –
autonomic nervous system
– ANS monitors heartbeat, body temperature, …
• Without any conscious effort on one’s part!
 Nature has evolved to cope with scale, complexity,
heterogeneity, dynamism and unpredictability, lack of
guarantees
– self configuring, self optimizing, self healing, self protecting, highly
decentralized, heterogeneous architectures that work !!!
 The goal of autonomic computing is to use appropriate
solutions based on current state/context/content based on
specified policies
Autonomic Computing
-6-
DP&NM Lab.
CSE, POSTECH
Autonomic Computing Attributes
 Self-managing systems that deliver
Increased
Responsiveness
Business Resiliency
Discover, diagnose,
and act to prevent
disruptions
Adapt to dynamically
changing
environments
Operational
Efficiency
Secure Information
and Resources
Tune resources and
balance workloads to
maximize use of IT
resources
Autonomic Computing
Anticipate, detect,
identify, and
protect against
attacks
-7-
DP&NM Lab.
CSE, POSTECH
Architecture concepts
 Autonomic computing system
– A computing system that senses its operating environment
– Models its behavior in that environment
– And takes action to change the environment or its behavior
Autonomic Computing
-8-
DP&NM Lab.
CSE, POSTECH
Architecture details
 Autonomic Manager
– Implementation that automates some management function and
externalizes this function according to the behavior defined by
management interfaces
Sensors
Effectors
Analyze
Monitor
Plan
Knowledge
Execute
Autonomic Manager
Data
Sensors
Effectors
Manageability
Interface
Resource
Manager
Autonomic Computing
Action
Managed Element
-9-
DP&NM Lab.
CSE, POSTECH
Architecture details
 Multiple autonomic managers
for multiple resources
– Autonomic managers are arranged
in hierarchical fashion
– Lower-level managers deal with
resources at smaller granularity
and/or smaller locality
Hierarchical
Autonomic Systems
 Top-level autonomic manager:
– Business decision-making
– Policy and service levels
Autonomic Computing
- 10 -
DP&NM Lab.
CSE, POSTECH
Architecture details
 Composition of Autonomic Systems
Workload
Manager
Arbiter
Planner
Provisioner
Registry
Reputation
Authority
Network
Broker
Server
Workload
Manager
Network
Sentinel
Event
Correlator
Monitor
Server
Database
Server
Negotiator
Aggregator
Sentinel
Database
Registry
Autonomic Computing
Broker
Monitor
Storage
Storage
Storage
- 11 -
Monitor
DP&NM Lab.
CSE, POSTECH
Architecture details
 Autonomic manager is a component that implements the
control loop
– Monitor Function
• the function that collects, aggregates, filters and reports
details (e.g. metrics, topologies)
– Analyze Function
• the function that models complex situations to understand
current system state.
– Plan Function
• the function that structures the actions needed to achieve
goals and objectives.
– Execute Function
• the function that changes the behavior of the managed
resource using effectors.
Autonomic Computing
- 12 -
DP&NM Lab.
CSE, POSTECH
IBM work
 Architectural blueprint
– Published at http://www.ibm.com/autonomic/pdfs/AC Blueprint White
Paper V7.pdf
 Autonomic Computing Manageable Resource Interface Specification
and Touchpoint Implementation Guide
– Published at
http://dl.alphaworks.ibm.com/techologies/aide/ManageabilityEndPpointSp
ecification0.51a_draft.pdf




550 patent applications
20+ specification to standards bodies
450+ self-managing autonomic features in 70+ IBM products
Services to accelerate customer implementation of Autonomic
Computing self-managing technologies such as problem
determination
 60+ business partners developing products according to Autonomic
Computing standards and technologies
 More than 100 customer references
Autonomic Computing
- 13 -
DP&NM Lab.
CSE, POSTECH
Application
 ANA Project ( 1/1/2006 ~ 12/31/2009 )
–
–
–
–
–
–
http://www.ana-project.org/
Autonomic Network Architecture – A European Union funded project in situated and autonomic communications (IST-FET)
Exploring novel ways of organizing and using networks beyond legacy internet technology
Ultimate goal is to design and develop a novel autonomic network architecture that enables
flexible, dynamic, and fully autonomous formation of network nodes as well as whole networks.
Scientific objective
• Architectural design that enables flexible, dynamic and fully autonomic formation of largescale networks in which the functionalities of each constituent network node are also
composed in an autonomic fashion
Technical Objective
• Building experimental autonomic network architecture, and to demonstrate the feasibility
of autonomic networking within the coming 4 years
Autonomic Computing
- 14 -
DP&NM Lab.
CSE, POSTECH
Application
 Autonomic Web Services
–
–
–
–
http://www.appliedautonomics.com
Identify and solve client problems quickly and accurately
Next Level of IBM Tivoli Storage Manager Solutions
Autonomic Services Framework: business data, work flows,
syndication services, and security
• Connect: collect business data from anywhere
• Understand: use work flows to simplify and understand
business data
• Communicate: publish the results for use
– Real examples
• How we save client’s time
• How we save client’s money
• How we provide value
Autonomic Computing
- 15 -
DP&NM Lab.
CSE, POSTECH
Application
 Enigmatec website
– http://www.enigmatec.net/
– Policy-Based Automation for the On Demand Enterprise
– EMS design Studio
• Intuitive drag-and-drop GUI
• Adaptable and re-usable automation designs
– Automation Pack
• Adapters, policies, and customized user interface
components for common domain-specific application
stacks and automation challenges
– EMS Operations Portal
• Intuitive, graphical, web-based monitoring and control
interface for administrators and management
• Real-time, in-flight view of all automated workflows
– EMS Platform
• Powerful service-oriented design
• Peer-to-peer run-time execution
• Vendor-neutral platform
Autonomic Computing
- 16 -
DP&NM Lab.
CSE, POSTECH
Application
 Handsfree Networks
– http://www.handsfreenetworks.com/
– ASI (Automated Support Infrastructure)
• Support and management offering that combines support and
management automation with complete device security
management
• Hide product complexity from the end user
• Make products highly reliable
• Insulate the user from the system management, support and
maintenance processes
– Practical autonomic computing
Autonomic Computing
- 17 -
DP&NM Lab.
CSE, POSTECH
Application
Autonomic Computing
- 18 -
DP&NM Lab.
CSE, POSTECH
Recent research result
 MACE 2006 (1st IEEE International Workshop on Modelling Autonomic Communication Environments)
–
–
–
–
–
–
http://www.manweek2006.org/mace/mace.php
Challenges, Principles and Concepts
• Challenges for Autonomic Network Management
• Emerging Principles of Autonomic Network Management
• On Concepts for Autonomic Communication Elements
• Autonomic Computing Design Methodology
Knowledge Engineering
• Ontology-based Engineering for Self-Managing Communications
• An OWL-S based architecture for self-optimizing multimedia over IP services
• Establishing workflow trust using provenance information
Promise Theory, Context and Situation-awareness
• Autonomic Pervasive Computing: A Smart Mall Scenario Using Promise Theory
• Onto-Context Manager Elements Supporting Autonomic Systems: Basis & Approach
• Situation-aware Composition of Low-level Traffic Handling Functions
Autonomic Computing and Autonomic Networks
• Autonomic Computing Approximated by Fixed-Point Promises
• Towards Autonomic Access Networks for Service QoE Optimization
• Towards Autonomic Control of Network Topologies
• Reconfiguration Discovery and Selection in the Context of Autonomic Management of
Cognitive Wirelss Infrastructure
Panel: Bio-inspired Network Management: a viable alternative or a pipe dream??
Autonomic Computing
- 19 -
DP&NM Lab.
CSE, POSTECH
Research Topic
 SASO 2007 (1st IEEE International Conference on Self-Adaptive and SelfOrganizing Systems)
–
http://www.labunix.uqam.ca/~jpmf/saso2007/
 Topics
– Self-* properties
• Self-oganization, self-adaptiveness, self-management, selfmonitoring, self-tuning, self-repair, self-configuration, etc
– Theories, frameworks and methods for self-* systems
– Management and control of self-* systems
– Robustness and dependability of self-* systems
– Engineering and control of emergent properties in self-* systems
– Biologically and socially inspired self-* systems
 Research communities
– Distributed artificial intelligence, networking, software engineering,
distributed systems, integrated management, robotics,
knowledge-based systems, machine learning, control theory,
mathematical optimization
Autonomic Computing
- 20 -
DP&NM Lab.
CSE, POSTECH
Research Issue & Challenges
Unprecedented Scale, Complexity, Dynamism, Heterogeneity,
Unreliability, Uncertainty
 Defining autonomic elements
– Programming paradigms and development models/frameworks
• Autonomic elements definition and construction
• Policies/constraints definition, representation, enforcement
 Constructing autonomic systems/applications
– Composition, coordination, interactions models and infrastructures
• Dynamic (policy/rule-based) configuration, execution and optimization
• Dynamic (opportunistic) interactions, coordination, negotiation
 Execution and management
– Runtime, middleware, infrastructure
• Discovery, interaction/coordination, communication, management, …
• Security, protection, fault tolerance, reliability, availability, …
Autonomic Computing
- 21 -
DP&NM Lab.
CSE, POSTECH
My work
 NGOMS (Next-Generation Operation & Management System)?
– http://ngoms.postech.ac.kr (not available)
– Autonomic Network and System Management Solution for Ubiquitous
Network (planning project, DP&NM Lab?)
– General Network and System Management Framework
• Supporting any type of Management Information Base
• Supporting any type of Management Protocols
• Supporting any type of Devices
• Service-Oriented, Scenario-Oriented, Business-Oriented Architecture
• Defining reusable features or components based on FODA method
• General network architecture and management design scheme
– Self-managed and self-adaptive system
• Context-aware, situation-aware
• Based on Autonomic management architecture
• Knowledge engineering ( co-work with experts )
– NGOMS Mobile ( 1st phase )
• Autonomic Mobile Device Management
• Autonomic Handover Management
• Self-diagnostics for mobile device ( on-going )
Autonomic Computing
- 22 -
DP&NM Lab.
CSE, POSTECH
Conclusion
 Summary
– Autonomic Computing
• Self-management system without human’s intervention
• Solution for large-scale and complexity of the distributed
systems
– Architecture
• Self-management: Self-configuration, Self-healing, Selfprotection, Self-optimization
– Autonomic Manager
• Managed resource
• Manageability interfaces – sensor, effector
• Monitor, analyze, plan, execute
– Applications
– Research Issues & Challenges
Autonomic Computing
- 23 -
DP&NM Lab.
CSE, POSTECH
References
[1] IBM Corporation: An architectural blueprint for autonomic computing.
White Paper, (2003)
[2] J.O. Kephart and D.M. Chess, "The Vision of Autonomic Computing,"
Computer, vol. 36, no. 1, Jan. 2003, pp. 41-52.
[3] R. Sterritt, M. Parashar, H. Tianfield and R. Unland, "A Concise
Introduction to Autonomic Computing," Journal of Advanced
Engineering Informatics, Engineering Applications of Artificial
Intelligence, Special Issue on Autonomic Computing and Automation,
Elsevier Publishers, Vol, 19, pp. 181 ~ 187, 2005.
[4] Wikipedia.org, http://en.wikipedia.org/wiki/Autonomic_computing/
[5] IBM Autonomic Computing Website, http://www.ibm.com/autonomic/
[6] IBM Corporation: Practical Autonomic Computing: Roadmap to Self
Managing Technology, January 2006
[7] Applied Autonomics, http://www.appliedautonomics.com/
[8] IPsoft, http://www-ip-soft.net/
[9] Enigmatec Corporation, http://www.enigmatec.net/
[10] HandsFree Networks, http://handsfreenetworks.com/
[11] Ana project, http://www.ana-project.org/
[12] MACE 2006, http://www.manweek2006.org/mace/mace.php
Autonomic Computing
- 24 -
DP&NM Lab.
CSE, POSTECH