| Popis: |
In situ analysis and visualization have been proposed in high-performance computing (HPC) to enable executing analysis tasks while a simulation is running, bypassing the parallel file system and avoiding the need for storing massive amounts of data. One aspect of in situ analysis that has not been extensively researched to date, however, is elasticity. Current in situ analysis frameworks use a fixed amount of resources and can hardly be scaled up or down dynamically throughout the simulation's run time as a response to changes in the requirements.In this paper, we present the challenges posed by elastic in situ analysis and visualization. We emphasize that elasticity can take various forms. We show the difficulties of supporting each form of elasticity with the state-of-the-art HPC technologies, and we suggest solutions to overcome these difficulties. The resulting four-way classification can be seen as a taxonomy for future elastic in situ analysis and visualization systems. |
