面向对象数据库中精细并发控制的自动化
Automating Fine Concurrency Control in Object-Oriented Databases
摘要：
为面向对象数据库提供适应性的并发控制提出了几个建议。然而，这些建议中的大多数都忽略了这样一个事实，即只考虑实例上的读写访问模式可能会导致比关系数据库中更少的并行性！本文解决了这个问题，优点很多：（1）方法的可交换性由编译器先验地和自动地确定，而没有可测量的开销，（2）可交换性的运行时检查与兼容性一样有效，（3）不需要为恢复指定逆操作，（4）该方案不排除更复杂的方法，并且，最后但并非最不重要的，（5）具有读写访问模式的关系型和面向对象的并发控制方案都包含在这个命题之下。
英文摘要：
Several propositions were done to provide adapted concurrency control to object-oriented databases. However, most of these proposals miss the fact that considering solely read and write access modes on instances may lead to less parallelism than in relational databases! This paper cope with that issue, and advantages are numerous: (1) commutativity of methods is determined a priori and automatically by the compiler, without measurable overhead, (2) run-time checking of commutativity is as efficient as for compatibility, (3) inverse operations need not be specified for recovery, (4) this scheme does not preclude more sophisticated approaches, and, last but not least, (5) relational and object-oriented concurrency control schemes with read and write access modes are subsumed under this proposition.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
