Analysis of big data processing technologies

Abstract

This article describes large amounts of information. Examples of large volumes are provided and some of their processing technologies have been analyzed. The advantages and disadvantages of the methods of processing parallel processing technologies of large volumes of data processing, the methods of selection of newly selected methods of selection and models are considered. These methods are compared with the distributed technology of data processing. This article discusses a few of the large volumes of data processing principles. They are: the principle of horizontal measurements, the principle of inverse stability, and the principles of data positioning. On the basis of horizontal measurements, any data processing system is expanding, with data doubling. In contrast to the principle of horizontal stability, the number of machines should not be increased. In the location of data, the data is distributed to a large number of machines. If the data is located on a server and processed in another, the cost of data transmission may exceed the cost of processing. Therefore, the most effective principle of data processing is the principle of information location processing. It is desirable to recycle data in a machine that will save them if possible. All models of large volumes of data processing are available in these three methods. This article analyzes a larger model of data that works with these principles. The MapReduce model also operates on the basis of the above three principles. This is also a model for distributed data processing to process large amounts of data in the computer cluster. In this article, several methods and models are analyzed and their advantages and disadvantages are discussed.