Section: Scientific Foundations
Optimized Protocol implementations and networking equipements
Participants : Jean-Patrick Gelas, Olivier Glück, Laurent Lefèvre, Pascale Vicat-Blanc Primet, Jean-Christophe Mignot, Ludovic Hablot, Sébastien Soudan, Fabienne Anhalt, Olivier Mornard.
The initial goal of the DARPA Internet Architecture was to develop an effective technique for multiplexed utilization of existing interconnected networks. Robustness was the first priority which strongly colored the design decisions within the Internet architecture. An architecture primarily for commercial deployment would have clearly placed the resource management at the beginning of the priority list. Some of the most significant problems with the Internet today relate to lack of sufficient tools for distributed management. For example, in the large Internet being currently operated, routing decisions need to be constrained by policies for resource usage. Today this can be done only in a very limited way, which requires manual setting of tables. This is error-prone and at the same time not sufficiently powerful. A key enabling factor of new services and protocols is then the ability for new software capabilities to self configure themselves over the network. Moreover, in the Future Internet, only trusting nodes should be able to communicate at will. Nodes should also be protected from nodes they do not want to communicate with. Virtualization, context-awareness and energy efficiency are then promising concepts for Future networks. However their potential and limits in the context of dynamic and self-organized high speed networks have to be studied.
Since several years, virtualization of the operating system is used in end system to improve security, isolation, reliability and flexibility of the environments. These mechanisms become a must in large scale distributed system. Virtualized resources is a new way of sharing in which group of users or activities (or trusting nodes) are given static shares, and only within these groups there is dynamic sharing. Virtual networks present ideal vantage point to monitor and control the underlying physical network and the applications running on the virtual trusting nodes. How virtualization can be also adapted and used in data transport networks and specifically in switching and routing equipments is an open question. For example, virtualization introduces an overhead which must be integrated to system performance models in order to forecast their behavior. Lot of performance problems on end systems but also on router's data plane have to be studied and solved to make the virtualization approach viable. Investigating these issues is one of the goals in this research axis.
On an other hand, the key enabling factor of new network services is programmability at every level; that is the ability for new software capabilities to self-configure themselves over the network. We explore the concept, "dynamic programming enablers" for dynamic service driven configuration of communication resources. Dynamic programming enablers apply to an executable service that is injected and activated into the network system elements to create the new functionality at runtime. The basic idea is to enable trusted parties (users, operators, and service providers) to activate management-specific service and network components into a specific platform. We study mechanisms and infrastructures required to support these components. We aim at providing new functionality to services using Internet facilities, addressing the self-management operations in differentiated and integrated services. The goal is the enhancement of the creation and the management (customization, delivery, execution and stop) of Internet services.
In this research axis we also explore the integration of context-awareness functionality to address two important issues : reliability of communications and energy consumption.
Session awareness : Most of the NGN services involve a session model based on multiple flows required for the signaling and for the data exchange, all along the session lifespan. New service-aware dependable systems are more than ever required. Challenges to these models include the client and server transparency, the low cost during failure free periods and the sustained performance during failures. Based on our previous work with FT R&D (“Procédés de gestion de sessions multi-flux”, N. Ayari, D. Barbaron, L. Lefèvre, France Telecom R&D Patent, June 2007), we continue to explore and propose session aware distributed network solutions which support the reliability mandatory to operators services (VOIP).
Energy awareness : Large scale distributed systems (and more generally next generation Internet) are facing infrastructures and energy limitations (use, cooling etc.). In the context of monitored and controlled energy usage, we plan to explore the proposal of energy aware equipments and frameworks, which allow users and middleware to efficiently use large scale distributed architectures.
We are developing solutions to dynamically monitor energy usage, inject this information as a resource in distributed systems and adapt existing jobs (OAR) and network (BDTS) schedulers to autonomically benefit from energy information in their scheduling decisions. This research is linked with experimental evaluation on Grid'5000 platform and inside the ALADDIN initiative.
