摘要： 分布实时数据库系统中事务间进行相互通信所产生大量的消息与日志，必然要延长事务的运行时间而影响系统的性能。该文通过对分布实时数据与事务特征进行分析，提出了一种数据分配算法，把数据合理地分配到相应的节点上，能有效地减少事务间相互通信的消息数量，从而提高了系统的性能。

关键词: 分布实时数据库；数据相关度；数据划分；数据存取价值

Abstract: In distributed real-time database systems, data and transactions are distributed in different sites. For ensuring consistency of systems, there must be communicated among the transactions which will generate a number of messages and logs, which increase the executing time of transactions. It is an important to real-time systems how to decrease the number of messages and logs among transactions. This paper proposes a data portioning policy by analyzing the data and transaction factors impacting system performance, which improves the performance of systems via reducing the intercommunication numbers among the transactions.

Key words: Distributed real-time databases; Data mutuality degree; Data partitioning; Data access priority