Research On Edge Cloud Computing Architecture And Data Migration Method

Cloud computing, integrating the computer technology and Internet technology,becomes a research focus as the strategic technology, which leads the innovation of thefuture information industry. The available commercial centralized platform of cloudcomputing has a drawback of high network load. Academia combines the concept ofcloud computing with distributed technologies and proposes the distributed cloudcomputing architecture. Independence of its each subsystem, the limited capacity ofproviding service, and the dispersion of service resources make it very difficult tomanage the resources in the distributed cloud computing architecture. With theseproblems, this paper improves this architecture and studies the data migration method.To these problems of distributed cloud computing that subsystems are independentof each other, service resources are scattered, etc. Based on the depth analysis of thecentralized and distributed architecture of cloud computing and reference the idea ofclassification management, this paper proposes the edge cloud computing systemarchitecture consisting of "edge cloud" and "core cloud", and also gives the definitionand function of "core cloud" and "edge cloud". With the help of the management andmonitoring of the core cloud to edge cloud, it decreases the network load of a singleedge cloud and solves the difficult problem of collaboration among distributed cloudcomputing architectures.In the edge cloud computing architecture, to guarantee the response time of cloudcomputing service and minimum the data redundancy of the system at the same time,this paper proposes and analyzes the data migration strategies of object-orientednetwork and pre-store in different conditions. On the basis of those strategies, this paperproposes a “data migration strategy based on the state of the network”, to flexiblycontrol the balance of the network traffic and storage and it guarantees the real-time ofservice response.Base on the edge cloud computing architecture, this paper analyzes thecharacteristic of resources in the edge cloud system, designs and implements theprototype system of edge cloud computing resources management, referencing the ideaof cascading architecture. After tested the system, the results show that it is capable ofmanaging the network and storage resources with the real-time needs of users, providers support for data migration with low real-time overhead.