PluriHis : A Highly Scalable Scheme of Distributed Historical Data Storage and Access
Overwhelmed by the massive volume of production data, traditional RDBMS-based historical storage solutions are going burst. A distributed data storage and access mechanism incorporating NoSQL designs is presented trying to solve the problem. With the orchestration of master nodes(2 machines, activestandby), real-time data incoming are routed to be cached on selected nodes to fulfill the load balancing requirements. Real-time data collected will pass two layers of local cache before finally dumped by each PluriHis node. Client modules on PluriHis nodes run in P2P mode and will actually register the name/location of a data point only after the corresponding request is successfully serviced. Master nodes will route all the data requests to the relevant PluriHis nodes and data query and retrieval is supported by a NFS-like network file sharing protocol. As Indicated by the benchmark results recorded, PluriHis is a highly capable candidate in implementing the next generation SCADA system with big data support.
Historical Data Production Data Storage Distributed Data Storage SCADA PluriHis
JIN SHU ZHANG YING BAI ZHONG-HE DONG SHENG-GANG
R&D Center,Nanjing SAC-Metso Control System Co.,Ltd
国际会议
西安
英文
1-5
2016-09-01（万方平台首次上网日期，不代表论文的发表时间）