Welcome Quality Management of Service Oriented Architectures 11th SoSORNet Workshop 22nd and 23rd Ap

1
WelcomeQuality Management of Service Oriented
Architectures11th SoSORNet Workshop 22nd and
23rd April 2008King's College London, Strand,
London

• Assel Akzhalova
• Connie Bau
• Nicolas Gold
• Iman Poernomo

2
QoS Management

• Service Level Management (SLM) is the process of
  managing the Quality of Service (QoS) demanded by
  clients and offered by providers.
• Has emerged as a key factor in enterprise
  systems.
• Traditional approaches to SLM are inadequate when
  dealing with complex SOAs
• Service-oriented architectures are compositional,
  dynamic and often distributed over the internet.
• For such architectures, SLM becomes a difficult
  problem that can no longer be handled by
  traditional monitoring/diagnostic tools.
• This is because of the dynamic, flexible,
  compositional and global natures of SOAs.

3
Purpose of this workshop

• This workshop is a forum for discussing the
  issues that are important to modern QoS
  management the monitoring of widely distributed
  components, dynamic adaptation strategies and the
  necessity for more sophisticated prediction and
  diagnostic analysis techniques.
• Aim to bring together researchers interested in
  the management of QoS in complex software
• Growing interest (a number of related workshops
  being held, including AQuSerM at EDOC)
• Lots of open questions
• Solutions cuts across research domains
  design-time QoS models, runtime verification,
  adaptive software and MDA, SOA, CBSE, Grid
  computing
• Encourage exchange of ideas from people based
  within these domains

4
A view of QoS Management
5
What we do at Kings (Palab)
Model Driven Management
Contracts Specification Negotiation Monitoring
Middleware infrastructures
Adaptation
Modelling
Diagnostics
Application domains Components, workflow, queued
communication networks
6
The Predictable Assembly Laboratory (Palab_at_Kings)

• Four objectives
• Model-driven generation of instrumentation and
  monitoring code.
• Model-based prediction of system behaviour at
  runtime.
• Runtime adaptation of architectures based on
  predictions, to pre-emptively avoid bad things
  happening.
• Guaranteeing model-transformations
• are correct

7
Our view of the world
8
Sessions

• Organization is approximate, as all research cuts
  across themes
• 2 future trends sessions dealing with ideas
  that have the potential to change the way we do
  QoS management
• Evolutionary computation
• Adaptive architectural models
• Session 1 Control and adaptation
• Session 2 Future trends -- potential for
  adaptive and evolutionary computation in QoS
  management.
• Session 3 Design and management
• Session 4 Middleware support
• Session 5 Model driven management
• Session 6 Future trends -- convergence of
  management with formal models of dynamic
  architectures

9
Format of workshop

• Interaction and discussion are key!
• Each session 20 minutes talks 10 mins
  questions
• At the end of the workshop, an open discussion
  will be held, about the papers we have heard, and
  how they might relate to one another (or to other
  work we know)
• Lets try to set the agenda for future QoS
  management research here and now
• As you listen to the talks today, note down
  topics/questions/themes/requirements for the
  final discussion session
• Sample topic should formal methods figure in QoS
  management? If so, where and how?

10
Follow-up workshop

• Guijun and Iman are running a workshop on
  Advances in Quality of Service Management
  (Aquserm) at EDOC 2008 in Munich
• Similar theme to this workshop, but with refereed
  proceedings
• Dates
• Submission 20 June 2008
• Workshop 18 September 2008
• Workshop will lead to a special issue of a
  leading journal

11
Enjoy the workshopand London