Architecture of highly available databases

Abstract

This paper describes the architectures that can be used to build highly available database management systems. We describe these architectures along two dimensions - process redundancy and data redundancy. Process redundancy refers to the management of redundant processes that can take over in case of a process or node failure. Data redundancy refers to the maintenance of multiple copies of the underlying data. We believe that the process and data redundancy models can be used to characterize most, if not all, highly available database management systems. © Springer-Verlag Berlin Heidelberg 2005.