Research Of Distributed Data Inconsistencies In The Cloud Computing Environment

Cloud computing is the traditional IT and communications technologies continueto blend and demand, and results in drive and promote business model. With the depthof cloud computing technology, more and more applications regard distributed filesystem as the underlying storage architecture, build database based on distributed filesystem become one of the current cloud computing management field direction, howto solve the inconsistencies of the data in distributed computing system for cloudcomputing become a key consideration.In a traditional centralized database, there have been effective SQL-basedtechniques for finding violations, however, with the changes in the structure of data aswell as the reality of the complexity of data semantics, these existing functionaldependencies are no longer applicable to distributed environment, distributed databasenot only bring a great challenge to data inconsistencies detection, but also requireshipping data from one site to another. In this thesis, we first propose a differentmethod of detection data inconsistencies, on the basis of the characteristics of thedistributed systems, through the analysis of a specific instance, and then using avariety of data transmission, so these physically dispersed, logically centralized datacan be processed, ultimately output inconsistent data, this centralized management,decentralized control structure greatly enhanced the parallelism and scalability of thesystem.This paper firstly introduces the background of cloud computing and distributedsystems, and then gives a definition of conditional functional dependencies, as well asthe role conditional functional dependencies play in the area of data cleaning, on thisbasis, we put forward the concept of extended conditional functional dependencies,and briefly discusses these distributed data, then divide the type of inconsistent databy storage, then introduce solution of detecting distributed data inconsistency one by one. The article focuses on solving the problem of distributed data transmission, thensimply describe the experimental conditions, finally, the last part is summary andoutlook.In order to verify the correctness and reliability of the scheme, we conducted aseries of experiments, to test the scalability of the system from the increase in thetuples of data and constraints change, and simulate the performance of a variety ofdata transmission in a distributed environment, and have achieve the desired results,these are fully illustrate this cloud computing environment for distributed dataprocessing program is feasible, this cloud computing distributed data processingenvironment has laid a good basis for future more in-depth study of cloud computingplatforms and distributed system.