支持海量数据操作的Oracle数据引擎的设计与实现  

Design and Implementation of Massive Data Access Engine of Oracle

作　　者：王喆[1] 潘懋[1] 戴婉薇[1] 郭艳军[1] 

机构地区：[1]北京大学地球与空间科学学院,北京100871

出　　处：《地理与地理信息科学》2008年第2期27-29,52,共4页Geography and Geo-Information Science

基　　金：国家自然科学基金项目"高精度三维复杂地质结构模型快速构建方法研究"(40742015)

摘　　要：三维地学建模多需海量数据支持,其数据库存储通常采用Oracle的LOB字段,而目前多数Oracle数据库访问接口无法处理大于4 G的LOB字段数据。该文基于面向对象思想,设计并实现了以OCI(Oracle Call Interface)技术为基础的Oracle数据引擎组件,将数据引擎操作的对象划分成多个层次,不仅能处理非海量的常见类型字段的数据,而且支持对4 G以上LOB字段数据的操作。该组件不仅突破了其他技术对Oracle LOB字段访问的限制,而且克服了直接使用OCI接口过于复杂、不利于复用等缺点,具有效率高、资源占用率低等优点。3D geography modeling is inseparable from massive data. The eommon method to store the massive data in database adopts Oracle LOB field. And most of the Oracle Database Access Interface is unable to deal with more than 4 G LOB field data. Based on the object--oriented design, this paper implemented the Oracle data access engine component which is based on OCI technology. The data objects which are operated by engine are divided into different levels. It will not only deal with the mass of the common data field types,but also support for 4 G or more data of LOB field over the operation. After the practice,it shows that the component is not only a technical breakthrough in the field of Oracle LOB visit restrictions. Moreover, it avoids the shortcomings such as complication of use and not conducive to reuse of OCI interface directly. It also has high efficiency, resources with low occuoancv rate.

关 键 词：ORACLE OCI 海量数据 数据引擎 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论] P208[自动化与计算机技术—计算机科学与技术]