| The primary purpose of Finance Data Cloud Platform-FDCP construction is to collecting financial information all around the world according to the national interest, and provide more comprehensive and accurate financial information service to national financial or non-financial institutions, facilitating them to participate in international market of Securities、foreign exchange、currency、fixed income、 commodities and energy. And then master data collection rights of financial information, enhance the authority of financial information releasing, form the right of financial product pricing, improve the speak right in the financial sector, effectively safeguard national financial security and information security.In March2006, Amazon introduced EC2service for the first time. Until today Cloud Computing can be summarily divided into Infrastructure as a Service, Platform as a Service and Software as a Service.Financial data can be divided into structured data and unstructured data. With the development of structured data and unstructured data, the boundaries began to blur, and often needs to be translated under certain conditions. Financial data has a huge amount and needs to be long-term preserved. FDCP faces the arduous difficulties in mass data storage, high demand of the real-timing and accuracy of financial data, and data distribution. Therefore, this paper mainly focus on the research and design of financial data distribution and storage on PaaS. Through the research of FIX protocol、NAS、SAN、Object Storage、Google GFS、HDFS, we mastered data storage and distribution technology.The data distribution on FDCP requires the flow control to be flexible to adapt to the data burst and effectiveness, and the storage methods suitable for off-site storage and off-site backup storage, meanwhile reflects the usability、security、performance and expansibility. The design of data distribution on FDCP contains five components: Data Communication、Data Delivery、Policy Controller、a series of control Policies and data service for the data storage. Data storage in FDCP is not only designed for data security, remote access, data consistency, but also managed to off-site backup and disaster recovery. Three levels of data storage (database storage, file storage, memory/cache storage) will be involved in this platform. With the technologies of backup virtualization, SOA, automation management, the platform is able to data off-site backup on FDCP. By automatic replicating and synchronizing among nodes, as well as remote volume replicating based on IP network on RAID, the platform is equipped with data-level disaster recovery and data consistency between two sites, which makes it can recover from off-site in disaster condition. With DNS access and access authority control layer, send the application requests forward to off-site application platform through application layer routing control to achieve remote access.In this paper, on research of the data distribution and storage on FDCP, we are committed to the design on future FDCP dealing with multi-types, multi-channel, multi-categories financial data. |
PDF Full Text Request