【作者】 廖国琼； 刘云生； 刘勇；

【Author】 Liao Guoqiong~(1,2),Liu Yunsheng~2,and Liu Yong~1 1(School of Cooperation Technology,Siemens Limited China,Beijing 100102) 2(School of Software,Huazhong University of Science & Technology,Wuhan 430074)

【机构】 西门子(中国)有限公司西门子中国研究院； 华中科技大学软件学院；

【摘要】 嵌入式实时数据库除需恢复数据库的内部一致性外,还需恢复时间一致性及复原物理世界(外部环境)的状态.在分析嵌入式实时数据库中数据和事务特征的基础上,讨论了嵌入式实时数据和事务的分类及其一致性约束.基于这些约束,给出了"识时"的恢复准则及策略.通过与传统恢复方法比较,所给出的策略在降低超截止期事务比率和减少系统拒绝服务时间等方面都具有较好的性能.更多还原

【Abstract】 In embedded database systems,apart from the internal or logical consistency,the time consistency and the states of external environment changed by them are also required to recover.In this paper,the classifications and consistency constraints of data and transactions in embedded real-time databases are given first.Then,time-cognizant recovery principles and strategies for different classes of data and transactions are discussed.It is shown on performance tests that the suggested recovery approach has better performances than traditional recovery techniques.更多还原