作者: Jesús Acosta-Elias , Leandro Navarro-Moldes
关键词:
摘要: One of the main challenges grid systems large scale and data intensive is that providing high availability performance, in spite unreliability delay occasioned by size Internet. Replication enables us to meet such a challenge with success. In context weak consistency, fast consistency algorithm prioritizes replicas demand. Nevertheless, only works well single zone demand, whereas multiple demand zones its performance poor. this paper, we propose an chosen according whereby each select leader subsequently construct logical topology, linking all together. way, changes are able reach without low forming barrier prevent from happening.