Experiences in Interactive Distributed Visualization

We report on our experiences in distributed interactive data visualization. The experiences are based on previously published work [1, 2, 3]. 1 Distributed Data Management We have designed and developed a distributed blackboard architecture for interactive scientific visualization. The architecture is centered around a blackboard and satellite processes that produce and consume data. Satellites implement the simulation, analysis programs, geometry mapping and rendering algorithms. The purpose of the blackboard is twofold. First, it acts as a shared data space for all satellite processes. Second, it acts as an event notification manager. Satellites can subscribe to events that represent state changes in the blackboard. Whenever such an event occurs, it will publish the event to all subscribing satellites. The, concepually centralized, blackboard is implemented as a distributed set of communicating blackboards. Many techniques from distributed computing are used to mantain data consistency, enhance performance, scalable data sets, etc. The blackboard architecture allowed us to address two important issues concerning interactive visualization environments. First, tight integration between simulation and visualization. Second, to allow qualitative and quantitative data analysis to operate simultanously a shared data space.