Research On Seismic Data Distributed Storage Strategy Based On Hadoop

With the rapid development of exploration area, the equipment for collecting seismic datakeep updating and the data size becomes larger. In order to improve the efficiency of processingthese data, varies methods are brought up by people. In fact, there are many factors influencing theefficiency when processing seismic data and they can be divided into two aspects: software andhardware, or to say, the configuration of access methods and access environment. However, thecontinuing renovation of those two factors is confronted with more and more difficulties andbrings enormous expense as well.To solve these two problems, this thesis is going to come up with distributed storagestrategies which can be used to optimize the access environment and improve utilization rate ofequipment, based on the characteristics of the storage of seismic data as well as the accesstechnologies of big data in view of Hadoop. In this thesis, the concrete research content is asfollows1.The combination of Hadoop distributed framework and seismic data characteristicsTo make adaptability research on data processing pattern of Hadoop distributed frameworkand relative storage and access properties of seismic data;to find out considerations on variousfactor need to be considered in seismic data distributed storage case; to make effectivecombination of Hadooop’s data access method and seismic data access method and put forwardthe overall framework of seismic data distributed storage strategy based on the premise oflow-cost cluster.2. Organizational strategy on seismic data distributed storage:According to the clustering characteristics of Hadoop clustering environment, reasonableconfiguration of seismic date was made between corresponding organization and environmentparameter, and make organization of distributed seismic data, to make its more effective storage inthe distributed file system of Hadoop. Make experiments to verify the most appropriateenvironment parameter configuration of seismic data characteristics and its optimal dataorganizational strategy.3.It designs seismic data functional module which is based onhadoop:To further verify the advantages of distributed calculation on seismic data of Hadoop, thisthesis shall describe the following procedures step by step: simultaneously develop Hadoop programming framework MapReduce and the current seismic data concurrent access; makecomparison of the function module between these two environments; verify the high efficiency ofHadoop seismic data of distributed storage by changing the corresponding environmentparameters; find out the influence of the changes between the number of distributed nodes and thesize of data on the efficiency of data access.Finally, to make conclusion on the research contents of this thesis, achieve their optimizationtechniques and put forward complete seismic data storage strategy. In this way, the correspondingoptimization techniques and its feasibility and effectiveness in this thesis can be verified.