Continuously adjusting the horizontal scaling of applications hosted by data centers appears as a good candidate to automatic control approaches allocating resources in closed-loop given their current workload. Despite several attempts, real applications of these techniques in cloud computing infrastructures face some difficulties. Some of them essentially turn back to the core concepts of automatic control: controllability, inertia of the controlled system, gain and stability. In this paper, considering our recent work to build a management framework dedicated to automatic resource allocation in virtualized applications, we attempt to identify from experiments the sources of instabilities in the controlled systems. As examples, we analyze two types of policies: threshold-based and reinforcement learning techniques to dynamically scale resources. The experiments show that both approaches are tricky and that trying to implement a controller without looking at the way the controlled system reacts to actions, both in time and in amplitude, is doomed to fail. We discuss both lessons learned from the experiments in terms of simple yet key points to build good resource management policies, and longer term issues on which we are currently working to manage contracts and reinforcement learning efficiently in cloud controllers.

This paper presents a 100% Java middleware that aims to offer an open and extensible foundation for building large-scale real-time networked virtual environment applications. This platform relies on a partial replication model in which the whole simulation space is spread on a federation of processes based on application-specific criteria. A configurable event communication framework allows arbitrary consistency and synchronization protocols to be implemented in close cooperation with the application semantics. This middleware has been experimented with success for multi-player game prototypes involving both human-driven and autonomous simulated entities.

The collision detection (CD) is an interesting feature which provides optimal performance to radio LANs such to make the latter comparable to wired LANs (e.g. Ethernet). The European HIPERLAN 20 Mbps standard is the natural environment for this implementation, since this standard involves the most powerful physical base and the most multivalent architecture for radio LANs. We describe one possible way to detect collision in a radio LAN, based on the so-called Comb strategy which is very similar to collision detection. This technique is developed and implemented in LAURA Esprit project.