Fault Tolerance Issues in MicroQoSCORBA 1

1
Fault Tolerance Issues in MicroQoSCORBA, a
QoS-Enabled andConfigurable Middleware Framework
for Embedded Systems
A. David McKinnon, Kevin E. Dorow, Tarana R.
Damania, David E. Bakken, Wesley E.
Lawrence July 2002
2
Future Middleware Directions

• Support wide ranges of (deeply) embedded H/W
• Resources vary widely (memory, power, etc.)
• Home appliances, Sensor networks
• Tailor to Application and H/W constraints
• Requires fine granularity of configuration
• Multi-property Quality of Service
• E.g., banking needs security and fault tolerance
• S/W Engineering tools for Embedded DOC
• Design patterns, Heuristics, Profiling metrics

3
MicroQoSCORBA Objectives

• Think Small, with fine granularity of
  configuration
• Tailor middleware to both application and
  hardware constraints
• Develop a multi-property QoS enabled MW framework
• Maintain CORBA interoperability
• Develop an IDL based framework, not just another
  customizable IIOP engine, that interoperates with
  TAO
• Create S/W Engineering tools that aid the
  developer
• Domain expert, but MW/QoS novice Contract
  employees

4
Lifecycle Time Epochs
5
Architecture Overview
6
QoS Fault Tolerance Taxonomy
7
MicroQoSCORBA Bottom Line

• MicroQoSCORBA the only framework that
• Is a bottom-up rethinking from the device level
  of what should be configurable, and in what ways
• Is tailorable for a given application to the wide
  range of
• Device constraints, and
• Application-dictated constraints
• with a fine granularity of configuration
  constraints
• Will support both functional and QoS properties
• Not just realtime, but security and fault
  tolerance are also key QoS properties

For more information visit http//microqoscorba.ee
cs.wsu.edu
8
Thank you
9
Key Questions

• Q1 Baseline Constraints
• What constraints must be imposed on middleware to
  achieve a small footprint (ignoring QoS for now)?
• Q2 QoS Properties
• What strength or flavor of QoS properties
  (security, fault tolerance, real-time) can
  reasonably be supported in resource-starved
  embedded systems?
• Q3 Network-Wide Composition
• What issues, optimizations, and tradeoffs must be
  made when a network of QoS-enabled devices
  integrated? How can we support global
  properties, metrics, and goals involving the
  entire distributed embedded system?
• Q4 Software Engineering Issues
• How can we use QoS aspects to make the middleware
  more manageable and quantifiable? How can we use
  patterns and tools to help the developer to be
  more productive?

10
Key Questions in Context