Research And Implementation Of Seismic Big Data Platform Based On Hadoop

With the development of the 13 th Five-year Plan,the Big Data has been upgraded to national strategic level,and has become the favorable opportunity and important starting point for industry transformation and promotion.As current focus,big data is the inevitable result for the development of information technology at present stage.Its strategic significance lies in the deep mining for huge data information,and provides us with powerful productivity through plenty of temporal and spatial dynamic data analysis.After nearly 10 years of development,seismic industry has built huge observation network,producing large amount of disciplines observation data such as premonition and seismic data every day,and has made us have the certain data base.It becomes particularly urgent to promote the seismic industry integrated data management and deep mining analysis service through big data technology.Research and implementation of seismic data platform is to explore the big data platform which is suitable for the characteristics of the data for the seismic industry and business needs,so as to meet future data acquisition,distributed storage,asset management,and so on.The seismic data platform provides reliable storage platform and efficient computing security for data mining and time and space dynamic analysis in the future.First of all,the paper expounds the research situation from two aspects of the project significance and technical environment,and combined with the characteristics of existing business system,it analyzes the current situation and introduces the purpose and significance of the research.Secondly,the technical architecture of seismic big data platform is described in detail,including: the overall overview of the Hadoop architecture,the composition and operation principle of the distributed file system(HDFS),the implementation mechanism and characteristics of the distributed computing(MapReduce),the implementation principle and the advantages and disadvantages of the column database HBase.Then according to the platform architecture,data collection,data reconstruction and data services four aspects,this paper expounds the design scheme of the platform.It focuses on the different data reconstruction methods according to different data sources and the encapsulated data interface for the convenience of data storage and calling.And it implements the basic data management,query and download,data visualization and so on.Finally,it describes the platform construction method,including hardware deployment,software configuration and reliability design,and describes and demonstrates the platform access control,query download and visualization module.Integrated the above content,the paper presents the ideas of design and architecture for seismic data platform,and realizes data collection,data reconstruction and data services 3 major business modules,9 functional sub modules.The seismic data platform realizes distributed storage,distributed search and data visualization and so on for seismic industry data tentatively.And it optimizes the deployment mode and reliability.And it provides the initial validation and basic platform for the seismic service,such as seismic data storage,deep mining,space-time cross analysis,data visualization and so on in the future.