Research Of Cloud Computing Based On Data Storage Technology

Rapid economic development has become the theme of social development, with the rapid development of Internet technology, the amount of information is in the explosive growth, people in the face of busy work, but also need to face a very large amount of data, more precisely. It requires a lot of space for storing data and store a large number of redundant copies of data. These data are also growing in a certain speed. Therefore, how to effectively save the data and how to quickly read data from a large number of the data, which will likely become a problem that businesses want to rapid development who need to considered. Compared with the traditional storage technologies, Cloud storage has its incomparable advantages: high reliability, versatility, scalability, high capacity. Business can use cloud storage improving efficiency and reducing costs.The thesis studies Research of Cloud Computing Based on Data Storage Technology. In this paper,we mainly to analysis two core components of Hadoop platform: hadoop distributed file system and MapReduce distributed computing model. After this ,we found some problem may exist in the Hadoop distributed file system, such as a single node exist performance bottlenecks and the NameNode poor scalability,after these analysis , we made a simple optimization to Hadoop distributed file system on this basis, Which can enhance the overall performance of the system and increase the capacity of the system processing data. In the final, we made a test for the system of improve before and after, experimental data show that improved system can reach we expect result in a certain loss we can accept. At same time, we study the popular Hadoop platform, we hope that people can learn more about Hadoop platform and promote its development in order to better solve our problem in the storage area.