[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
[stp-pmc] First thoughts about a Component based SOA stack based on the Fractal reflective component model
|
Title: Message
Dear
all,
You'll find
hereunder some recent notes proposed by INRIA, summarizing ideas and thoughts
about what could be a future Component based SOA stack, based on the Fractal
reflective component model.
Have an interesting
reading, and please feel free to react, comment and propose.
Best
regards,
Alain
--
Alain
BOULZE
INRIA Rhone-Alpes
- ObjectWeb
SOA Project
Coordinator
email :
alain.boulze at objectweb.org
cell : +33
(0)6.21.09.43.66
phone : +33
(0)4.76.61.54.65
The vision of middleware that underlies several efforts in
ObjectWeb and in INRIA research work is that of the construction of software
infrastructures using a reflective component model (as currently exemplified by
the Fractal component model).
This approach provides several benefits:
- first, those that accrue with making explicit the software
architecture of a complex system, in terms of reduction of architetural
erosion, better maintenance and increased reusability of system parts
- second, immediate gains in terms of system configurability
and administrability, as demonstrated e.g. by the use of Fractal for the
construction or re-engineering of
several ObjectWeb middleware components:
Speedo, Joram on Dream, Jonas a la Carte, Petals
- third, it provides the
means (as demonstrated by the AOKell and - to a lesser extent - Julia implementations of Fractal to bring
together software architecture-based middleware design and a proper separation
of non-functional aspects made possible by techniques such as Aspect-Oriented
Programming
This vision could lead some works in order to develop a new
generation of middleware components for the ObjectWeb code base, providing in
particular enough foundations for the assembly of a complete, highly modular,
standards compliant "SOA stack", as well as a component-based re-engineeering of
major ObjectWeb code base elements such as JOnAS and JORAM.
It could help
too to evolve the ObjectWeb Fractal
component technology towards a model providing a full-fledge integration
of AOP and components, with linguistic support (e.g. ADL) for dynamically
evolving architectures, and the support of architecture
constraints.
Distributed systems management
The next frontier for systems and middleware research is
probably the construction of self-managed distributed systems, including notably
automating the management of system configurations, performance, failure
handling, and security. In the ObjectWeb community, work at INRIA on Jade, for instance, provides an early attempt at
developing a component-based framework for self-managing systems. The goal of a
work on this theme would be twofold :
- to develop a comprehensive system architecture for
distributed systems self-management, together with the necessary software
technology tools (e.g. in terms of ADL extensions, policy languages and
suppoting frameworks)
- to develop control technology (algorithms, and tools - e.g.
exploiting planification and statistical techniques) required for the
construction of self-managing capabilities.
ESB domain
Some studies have already
started around the development of the JBI ESB Petals project, and are focusing
on designing the functions and the behaviour of the Router, structuring the
Router for a distributed environment, and representaing the ServiceAssemblies,
using Fractal.
Cooperation with OrientWare
On those
themes, experience gained and approaches developed by our OrientWare partner, e.g. at the U. of Bejing around ABC component
model and the management of PKUAS, at the National Defense U. around CCM and at
ISCAS on QoS management in J2EE servers, should prove of particular
interest.
