一种海量数据下的Lucene全文检索解决方案

A Solution for Lucene-based Full-text Search under Massive Data Environment

针对海量数据下的全文检索遇到的索引时间过长、I/O过高的时间较长、检索响应时间较长的问题进行了分析。提出了一种以索引类型分解、索引分块、多索引联合检索以及使用RMI提供远程检索服务相结合的解决方案。实验和生产环境中的应用表明,此方案能解决上述问题并能够提供稳定、高效的搜索服务。

This paper analyzed the bottlenecks of full-text information retrieval under the massive data environment：index-building takes too much time、I/O rates stand at a high level for a long time、response time is too long.Propose a index organize method which based on data organized by the its type、multi-indices、multi-indices joint searching and using RMI technology to provide search services solution.Experiments and the application under the production environment,this program can address these bottlenecks and to provide a stable and efficient search service.