| |
|
|
|
|
|
|
Overview
Motivations and Goals
Architecture
Publications
Overview
Rudder is a coordination middleware with
intelligent deductive capabilities for Autonomic
Applications. Rudder adopts multi-agent based
management and reactive tuple space coordination
technology to provide a coordination environment to
develop and enable self-managing systems. The goal of Rudder is
to harmoniously integrate the multiple individual
autonomic component activities to the whole meaningful
system behaviors adapting to the changing requirements
and run time context.
Motivations and Goals
Wide-area
distributed computing environments have enabled a new
generation applications. Autonomic applications capable
of managing themselves using high-level policies can
effectively deal with complexity, dynamism, and
heterogeneity of the underlying Grid. Achieving
autonomic self-managing requires middleware support for
context-awareness, self-awareness, knowledge-based
analysis and planning, plan selection and execution. The
goals of Rudder:
-
Enable dynamic
and opportunistic interactions of components and
services.
-
Enable
applications to be dynamically composed and
configured to adapt to on the current requirements
and environment.
-
Provide
mechanisms for dynamically defining, configuring,
routing and executing policies and rules.
TOP
Architecture
Rudder employs a context-aware agents
framework together with a coordination model based on a
global decentralized tuple space. Self-adapting
autonomous agents define adaptive strategies and
policies, while the reactive tuple space enables polices
distribute deployment and execution to effectively
support self-managing autonomic applications.
Agent Framework
Software agents that can dynamically define, deploy and
execute rules to achieve their self-adaptive behaviors,
operate within flexible organizational structures while
being situated in the dynamic environment context.
Agents communicate, negotiate and coordinate with each
other by associatively reading, writing or removing
tuples using different pattern-matching algorithm, which
suits well for Grid applications to discover and
exchange information in the uncertain, dynamic and
heterogeneous environment. The agent framework consists
of three types agents:
Component Agent
System Agent
Composition Agent
Decentralized
Reactive Tuple Space
A decentralized
reactive tuple space can effectively and scalably
supports the distributed agent coordination and enable
the self-managing properties of the autonomic
application. Rudder coordination space is assumed to be
a global abstraction with dynamic programmable reactive
behaviors. It can be fully programmed by administrators
and peer agents to enable environment-specific and
application-specific coordination, thus providing more
secure and flexible coordination activities. It
implemented with distributed and flexible matching
engine and reactive semantics, and builds on a scalable
peer-to-peer content-based routing engine on top of the
Chord Overlay to implement semantically searchable
distributed hash table. It can provide:
-
Coordination
service for distributed agents
-
Mechanisms for
rule definition, deployment and enforcement.
-
Runtime adaptive
polices can definition and executed
Publications
Z. Li
and M. Parashar, "Rudder: A rule-based multi-agent
infrastructure for supporting autonomic grid applications",
In Proceedings of the International Conference on Autonomic
Computing, New York, NY, 2004.
Z. Li,
H. Liu and Parashar, "Enabling Autonomic, Self-managing Grid
Applications", accepted by International Workshop on Self-*
Properties in Complex Information White Paper, Italy 2004.
TOP
|
| |
|