Distribution Of Data Blocks And Optimization Of Sever Utility Problems Research In Cloud Storage

The concept of cloud storage appears with the development of the cloud computing. This paper researches the cloud storage mechanisms, and analysis the cloud storage system data storage and data access.In the cloud storage environment, data is backed up by more than the fixed size data blocks are stored in the cloud cluster server, and all the data blocks are stored in the form of a document under the Linux environment. System has a central master node, all of the data block operations are coordinated through it. The main work of this paper include:Cloud storage environment through the network topology and location information based on the relationship between the server location analysis, collection of data blocks in the cloud environment, hierarchical network topology map. Through the collection of data block distribution analysis of the structure proposed in the cloud storage environment risk assessment model of storage resources, and gives specific examples of risk calculation algorithms and analyzes the environment in the cloud storage resources based on risk research Provides a theoretical basis.Cloud storage environment for modeling, formal resources and server resources block some of the properties, set partitioning problem of the importance of resource distribution and the set partitioning problem is given using the variance of goals to measure the mathematical proofs. Cloud data storage system through the resource properties of the block, on the cloud storage system, the distribution of the data block resource model mathematical problems, and propose cluster server performance based on cloud data block distribution algorithm LPT-IGA. LPT-IGA algorithm is based on the LPT algorithm for improved genetic algorithm. Firstly, improved genetic algorithm initial population generation, the use of many of the LPT algorithm to generate the problem solution, and then use the fitness value-based roulette wheel selection algorithm selection by genetic algorithm initial population; Secondly, to improve coding, all coding Data blocks in accordance with the direction of resources are being accessed frequency (heat) descending direction coding; Finally, the proposed crossover operation and mutation encoding a random direction in accordance with the improvement increases the probability of crossover and mutation methods Operation.Data blocks for the cloud cluster server performance optimization, the first among the cloud cluster environment to maximize server performance issues of mathematical modeling, in this mathematical model based on the given approximation algorithm FFD-LPT. FFD-LPT algorithm DLPT improved binary search algorithm is the upper bound of the algorithm the search space smaller.