摘要
在不影响hadoop分布式文件系统分散式存储结构的前提下,结合动态副本存储和伽罗华有限域理论对范德蒙码的计算方法和计算模式进行整体优化,降低了编译码的时间代价和计算的内存压力,节约了hadoop分布式文件系统约35%的存储开销,提高了hadoop分布式文件系统节点负载均衡能力和译码恢复效率。这种算法更适合于医疗专业文书的处理,解决临床科研需求和数据供给2个方面的问题。既能节省了存储容量,可以容纳与日俱增且越发复杂化的医疗数据,又能降低硬件服务器成本,为医院节省资金开销,更能快速查询获取数据池中的有效数据,让这些躺着的数据变成活的,充分发挥他们的临床使用价值和科研价值。这一套完整的、系统的优化方案,为未来hadoop分布式文件系统的发展提供了一条有效途径。
On the premise of not affecting the HDFS decentralized storage structure, the dynamic copy storage and galohua finite field theory was combined to optimize the calculation and calculation mode of van der Monde code, so that the time cost and the memory pressure of the coding and decoding were reduced. About 35% storage costs of HDFS were saved, and the balance ability of the node load and decoding efficiency of the HDFS system were improved. This algorithm is more suitable for the process of medical professional documents, and meets clinical research needs and data supply. It can save the storage capacity, can accommodate the increasing and more complex medical data, can reduce the cost of hardware server, save the capital cost for the hospital, quickly query and obtain the effective data in the data pool, make the lying data live, and give full play to their clinical use value and scientific research value. This complete and systematic optimization plan provides an effective way for the development of HDFS in the future.
作者
杨莲
郭良君
马磊
王圣芳
Yang Lian;Guo Liangjun;Ma Lei;Wang Shengfang(Shandong Institute of Cancer Prevention and Control,Jinan 250117,China;Jinan Children's Hospital)
出处
《中国医院统计》
2019年第1期75-78,共4页
Chinese Journal of Hospital Statistics
基金
山东省医学科学院院级科技计划项目青年基金(2016-30)
