Abstract: This paper proposes the method of large file storage by blocking partition based on Hadoop and massive agricultural resources retrieval method, and establishes the massive agricultural data management platform based on Hadoop. Test and practice results show that this platform provides a technical direction for the realization of massive, distributed, heterogeneous unified agricultural knowledge resources, efficient organization and management.

Key words: massive agricultural data, management platform, distributed storage, meta data, resource retrieval

摘要： 使用传统分布式数据库架构存储和管理海量农业数据存在资源效率不高及存储能力不足的问题，为此，在Hadoop的基础上研究海量农业数据资源组织存储与检索技术，提出基于Hadoop的大文件分块存储方法和海量农业数据资源检索方法。测试及实践结果表明，该方法为构建海量农业数据资源管理平台提供了支持，实现了海量农业数据资源高效的组织和管理。

关键词: 海量农业数据, 管理平台, 分布式存储, 元数据, 资源检索

CLC Number: 

• TP391