|
European Research Network on Foundations, Software Infrastructures and Applications for large scale distributed, GRID and Peer-to-Peer Technologies |
|
Research group on Advanced programming models
Goal
The activities of this research group are all related to the coordination of the research activities of the Programming model Institute partners on advanced programming models that can be built on top of, or integrated with, or usefully exploited in conjunction of the component based programming model which will eventually be defined within the Institute.
In particular, within this task, several research tracks will be coordinated, concerning:
- ``theoretical'' computation models that can be used to support the grid component model,
- performance models developed to cover major grid computing parallel patterns,
- program transformation techniques, aimed improving the performance features of the user supplied code via automatic program rewriting techniques
- autonomic component and parallel application control, taking care of all the details related with the management of component interaction with the underlying grid middleware,
- advanced security issued related, as an example, to secure component code deployment, secure data transmission, authenticated component access, etc.
- ``invisible grid'' concept, as defined in the NGG expert group documents and provided to the application programmer level, thus allowing efficient grid application development without actually requiring any kind of knowledge in specific, grid related programming techniques.
The expected outcome of this research group is both related to the grid
component model whose research activities are coordinated in Task 3.2
and to the research activities of other CoreGRID Institutes. In
particular, Task 3.3 being related to ``advanced'' programming models,
its activities are very close to the ones performed in the Institute
on Problem solving environments, tools and GRID systems. This because
both the ``advanced programming environments'' simply built on top of
the grid component model of Task 3.2 and the integrated programming
environments that are investigated in the Institute on Problem solving
environments, tools and GRID systems aim at the very same goal:
raise the level of abstraction provided to the grid
application programmer.
The activities related to this research group build on some experiences of
Institute partners that either pre-existing the CoreGRID activities or
that have been (and possibly are currently being) developed within
``private'' (that is non-CoreGRID) partner research projects. Such experiences include:
- the ASSIST structured, high performance programming environment,
supporting component programming and providing user with components
including autonomic managers taking care of ensuring (best effort)
the user supplied performance contracts (aka QoS requirements) and
with components modeling common parallelism exploitation patterns,
- the HOC programming environment, providing the user with grid
programming components, which are pre-packaged with parallel
implementations and middleware support and accessible via Web
services, abstracting over the middleware platform.
- the experience concerning the usage of semantically well founded
and elegant abstract programming models, such as the ORC model by
Misra.
The activities related to Task 3.3 will concern the integration of the existing, related research activities at the partner institutions as well as the development of new research activities either financed by new projects or added to existing ones.
Partner of the Institute will exploit short visit and researcher exchange programs.
Participants
Group | People |
UNIPI | Danelutto, Vanneschi, Aldinucci, Coppola, Campa, Bertolli, Veraldi |
UNIPASSAU | Lengauer, Groesslinger |
INRIA | Caromel, Henrio, Perez, Andre, Buisson |
WWU Muenster | Gorlatch, Duenwebber |
VUA | Kielmann |
QUB | Kilpatrick, Clint |
UPC | Gabarro |
Documents
Short visits
Links
|