返回Google圖書搜尋

Dynamic Load Balancing for Active Objects on Computer Grids

出版	2006
URL	http://books.google.com.hk/books?id=KTCQygAACAAJ&hl=&source=gbs_api

註釋This thesis sets the foundations for the development of load-balancing algorithms for the active-objects model defined by Proactive in the context of large-scale networks (grids). Proactive is an open-source Java middleware which implements the active-objects paradigm : each active object has its own control-thread and can independently decide which order to serve incoming method calls. To add efficiency to the active objects paradigm, Proactive provides a migration mechanism, but the migration operation comes with a communication penalty : an active object must migrate with its complete state. Therefore, proactive applications are sensitive to latency. When several active objects with identical functionality are deployed, a load-balancing algorithm can be used to improve the performance of an application using that functionality. Hence we developed and studied a load-balancing algorithm for active-objects which belong to a parallel application. This algorithm was validated using simulations and introducing our model of grids. We chose to perform the validation in the context of a large-scale network, because the environment where the active objects run is usually composed of multiple clusters of resources. Our grid model is based on observations and measurements of what we consider key characteristics for active-objects load-balancing : processing capacity and inter-resource communication latency.