Capita Selecta Distributed Systems

1
Capita SelectaDistributed Systems

• Module 4
• Autonomic Computing
• Tom HolvoetDanny Weyns

2
Introduction / presentation

• Who is who
• Capita selecta - distributed systems
• Part I Aspect-oriented middleware (AOM)
• Wouter Joosen, Eddy Truyen
• Part II Distributed algorithms
• P. Verbaeten
• Part III Study of context-aware computing
  systems
• Yolande Berbers, Davy Preuveneers
• Part IV Autonomic computing
• Tom Holvoet, Danny Weyns

3
Overview

• Goals
• Set-up
• Planning
• Evaluation
• Arrangements

4
Goals

• Autonomic computing (AC)
• what?
• requirements for self-management
• linking AC with software architecture
• architecture-based approach for AC
• requirements for architectural support for AC
• component model
• Software architecture
• importance / potential
• Research experience
• paper study (not paper reading)
• paper assessment
• critical reflection

5
Module set-up

• Reading club
• in-depth study of (limited amount of) material
• Session preparation
• study (not just read) of session papers /
  material
• all 5-6 issues for discussion / position
  statements
• crucial!
• may be provocative
• Session
• brief introduction
• discussion based on proposed discussion
  issues / position statements
• Adopt a critical attitude
• Put in perspective
• terminology not always consistent in material
• time frame
• authors, institution

6
Planning

• 3 reading club sessions
• Session 1 problem statement / context (Mar 25)
• autonomic computing introduction
• AC as an architectural challenge
• Session 2 (Apr 1)
• Session 3 (Apr 22)

7

• An Architectural Blueprint for Autonomic
  Computing, IBM, Autonomic Computing White Paper
  4th Edition, 2006
• Self-Managed Systems an Architectural Challenge,
  J. Kramer, and J. Magee, Future of Software
  Engineering, FOSE 2007
• Endogenous versus Exogenous Self-Management, D.
  Weyns, R. Haesevoets, B. Van Eylen, T. Holvoet,
  and W. Joosen, Software Engineering for Adaptive
  and Self-managing Systems, SEAMS 2008
• Reflecting on Self-Adaptive Software Systems, J.
  Andersson, R. de Lemos, S. Malek, D. Weyns,
  Software Engineering for Adaptive and
  Self-Managing Systems, SEAMS 2009

8
Evaluation

• Factors
• understanding material
• critical reflection
• attitude
• ... demonstrated through
• preparation material
• discussion during reading club session
• (if necessary individual discussion afterwards)

9
Practical arrangements

• At end of each session
• Sneak peak next week
• Preparation material
• E-mail day before session, before 6PM!
• what
• for all participants
• 5-6 discussion issues / position statements (1/2
  - 1 page max)
• for those responsible for introducing a paper
• brief introduction slides
• to
• Tom Holvoet and Danny Weyns
• Task assignment brief paper intro
• 4-5 slides max.
• dont present paper
• present key issues - we all read the paper
• starting point for discussion
• Suggestions
• later

10
(No Transcript)
11
Sneak peakSession 1Problem statement Context
12
Problem statement Context

• Four papers on
• Problems underlying the need for self-managing
  systems
• Visions on how to tackle the problems
• Papers
• Autonomic Computing IBMs Perspective on the
  State of Information Technology, Paul Horn, IBM
  research, 2001
• Self-Managed Systems an Architectural Challenge,
  Jeff Kramer and Jeff Magee, Future of Software
  Engineering 2007
• Endogenous versus Exogenous Self-Management, D.
  Weyns, R. Haesevoets, B. Van Eylen, T. Holvoet,
  and W. Joosen, Software Engineering for Adaptive
  and Self-managing Systems, SEAMS 2008
• Reflecting on Self-Adaptive Software Systems, J.
  Andersson, R. de Lemos, S. Malek, D. Weyns,
  Software Engineering for Adaptive and
  Self-Managing Systems, SEAMS 2009
• Study in this order

13
Problem statement context

• IBM
• Significant research lab that studies strategies
  for building self-managing systems
• Paper
• Explains the growing complexity problems with IT
• Puts forward the idea of autonomic computing and
  explains key elements of autonomic computing
  systems
• Commercial point of view

14
Problem statement Context

• Kramer and Magee
• Imperial College London
• Renowned researchers in the field of software
  engineering/architecture
• Paper
• Architectural model for self-management based on
  architecture for robotic systems
• Research issues
• Potential to become reference paper in research
  community

15
Problem statement Context

• D. Weyns, R. Haesevoets, B. Van Eylen, T.
  Holvoet, and W. Joosen
• Software Engineering for Adaptive and
  Self-managing Systems, SEAMS 2008
• KULeuven-DistriNet
• Distributed software software architecture
  multi-agent systems
• Paper
• Differentiates two types of autonomic systems
• Self-organizing (endogeneous) vs.
  Architecture-based (exogeneous)

16
Problem statement Context

• J. Andersson, R. de Lemos, S. Malek, D. Weyns
• Software Engineering for Adaptive and
  Self-managing Systems, SEAMS 2009
• KULeuven-DistriNet
• Distributed software software architecture
  multi-agent systems
• Paper
• puts autonomic computing in perspective of
  reflection

17
Assignment

• An Architectural Blueprint for Autonomic
  Computing, IBM, Autonomic Computing White Paper
  4th Edition, 2006
• Jef Maerien
• Self-Managed Systems an Architectural Challenge,
  J. Kramer, and J. Magee, Future of Software
  Engineering, FOSE 2007
• Kenneth Piron
• Endogenous versus Exogenous Self-Management, D.
  Weyns, R. Haesevoets, B. Van Eylen, T. Holvoet,
  and W. Joosen, Software Engineering for Adaptive
  and Self-managing Systems, SEAMS 2008
• Rob Vanden Broucke
• Reflecting on Self-Adaptive Software Systems, J.
  Andersson, R. de Lemos, S. Malek, D. Weyns,
  Software Engineering for Adaptive and
  Self-Managing Systems, SEAMS 2009