A Case for Staged Database Systems

Traditional database system architectures face a rapidly evolving operating environment, where millions of users store and access terabytes of data. In order to cope with increasing demands for performance, high-end DBMS employ parallel processing techniques coupled with a plethora of sophisticated features. However, the widely adopted, work-centric, thread-parallel execution model entails several shortcomings that limit server performance when executing workloads with changing requirements. Moreover, the monolithic approach in DBMS software has lead to complex and difficult to extend designs. This paper introduces a staged design for high-performance, evolvable DBMS that are easy to tune and maintain. We propose to break the database system into modules and to encapsulate them into self-contained stages connected to each other through queues. The staged, data-centric design remedies the weaknesses of modern DBMS by providing solutions at both a hardware and a software engineering level.