A distributed cache scheduling system and method based on social relations

Abstract

The invention relates to a distributed cache scheduling system based on social relationships, including a proxy server installed in a network communication system, including: a social relationship calculation module, a clustering module, a cache server management module, a request management module and a communication module; The social relationship calculation module calculates the social relationships between users based on the social information of the users; the clustering module clusters the social relationships of the users, and clusters the users with similar social relationships into the same social group; the request management module is used for Create requests, send requests, and process requests; the cache server management module is used to manage the resources stored in the local cache server, and to search for content or add tags according to requirements; the communication module is used to realize the communication between the proxy server and users Communication, communication between proxy servers, and communication between proxy servers and cache servers.

Description

technical field [0001] The invention relates to content distribution technology in the field of network communication, in particular to a distributed cache scheduling system and method based on social relationships. Background technique [0002] In recent years, with the increase in the number of computers and the popularity of devices such as smartphones and mobile tablets, network traffic has exploded. According to Cisco's official statistics, the global IP traffic has increased 8 times in the past 5 years; it is estimated that in 2016, the global IP traffic will reach 1.3 Zeta bytes (ZB). The huge traffic makes the current network face many problems such as bandwidth shortage and low network transmission efficiency. Therefore, caching technology is introduced to alleviate the traffic pressure problem faced by the existing network. [0003] There are two main types of caching technologies: centralized caching and distributed caching. When the centralized cache system pr...

AI Technical Summary

Problems solved by technology

However, the traditional cache scheduling technology is still an extensive technology. It is difficult to meet the repetitive, complex and personalized needs of users by only relying on the local characteristics of local users to predict future content needs.

Scholars at home and abroad have proposed many optimization schemes, but these schemes still have the following shortcomings: 1) It is difficult to completely solve the above-mentioned problems by relying on local features to optimize the network; 2) Some methods only consider the similarity of users' social attributes to calculate 3) Most methods only consider the improvement of user service quality, but do not take into account the load capacity of the server

The disadvantages of this type of method include: there is no optimization for the scheduling between clusters and the content distribution of the whole network; the resources requested by users may need to be scheduled from cache servers other than the local, making this method difficult for user services The improvement in quality is not obvious

This solution has the following deficiencies: (1) The optimization goal of this solution is the transmission of information between users and the sharing of content among users, and the content provided by network operators is not included; (2) This solution is more inclined to deal with and store large data. When users need network content, they use the method of extracting from the cache storage of "friends" instead of distributing the data to the local cache server

The disadvantages of this type of system include: the central management system undertakes a large amount of computing and communication work. With the growth of network content and the increasing number of cache servers, the load of the central management system is also increasing, resulting in the overload of the central control system. Heavy; and the central management system is overloaded or even collapsed, which will cause the entire network to fail

Therefore, this type of architecture brings great hidden dangers to network stability.

Images