Research Internships 2005-2006
General context
The general context of this framework is components models. Our
team is highly involved in the design of a Grid Component Model (GCM),
which consists in a Grid-oriented extension of the Fractal component
model.
In general in component programming, but even more specifically in
distributed and Grid computing, components need to be highly
adaptative, a great part of adaptativeness relies on reconfiguration of
component systems, i.e. dynamically (during execution) bind, unbind,
add and remove components.
Fractal defines "basic/predefined" controllers that should be present
in most of Fractal components. We are interested here in binding,
life-cycle, and content controllers (they allow to bind/unbind
components, start/stop components, add/remove components inside another
one). The primitives proposed in Fractal are sufficient for expressing
any reconfiguration, but they are rather low-level.
The Oasis team is developping a distributed implementation of Fractal
(over ProActive), together with model checking tools adapted to those
components systems
Objectives
The objective of this internship is to design a higher level of
reconfiguration primitives allowing to achieve complex operations, but
also to trigger such operations on specific events.
This internship consists in designing a set of high-level primitives
(e.g., replace, add and bind, unbind and remove, duplicate, recursively
add, ...) for reconfiguration ensuring more correctness properties than
the Fractal ones, and more autonomicity. By providing higher level of
primitives, we want to help the programmer to design safe scenario. For
example a replacement seems safer and easier to verify than the
equivalent sequence (stop+unbind+remove+add+bind+start) that simulates
it in Fractal.
From this point, several objectives can be envisioned:
- implement a controller for these primitives
- verify/prove the correct behaviour of the application of these primitives
- design a way of specifying synchonisation between those primitives and the application
- apply model checking techniques to prove the correctness of the behaviour of a given example, using the new reconfiguration primitives, and perhaps the synchonisation features
- study how those reconfiguration, especially with synchronisation, can be adapted to "shortcuts" (a shortcut is a direct linnk between components, bypassing several intermediate membranes)
The objective of the internship is not necessarily to fulfill all the preceding goals.
Potential advisors : Ludovic Henrio and Eric Madelaine
and Francoise Baude
Phone : 04 92 38 71 64 Email : Ludovic.Henrio@inria.fr
Team : OASIS (INRIA Sophia Antipolis -- I3S --
CNRS)
Prerequisite : Some knowledge in Object-oriented languages, distributed programming, formal models.
Internship location:
Sophia Antipolis, between Nice and Cannes,
France
A few internal references
- http://www-sop.inria.fr/oasis/ProActive/
- Verification of Distributed Hierarchical Components Tomás Barros, Ludovic Henrio, Eric Madelaine- FACS 2005
- Behavioural Models for Hierarchical Components Tomás Barros, Ludovic Henrio, Eric Madelaine- SPIN 2005
- From Distributed Objects to Hierarchical Grid Components Francoise Baude and Denis Caromel and Matthieu Morel International Symposium on Distributed Objects and Applications (DOA), Catania, Sicily, Italy, 3-7 November 2003