Research On Efficient Container Supply And Reliablility Safeguard Technology In Edge Computing

With the booming development of Internet of Things technology and the gradual popularization of the mobile 5G network,the number and scale of edge devices are growing rapidly,and massive real-time data are also generated.The traditional cloud computing paradigm is unable to meet the requirements of low latency and high availability of edge device applications.Therefore,edge computing has been widely used as the expansion and supplement of cloud computing.In addition,the deep combination of edge computing and container technology provides a good solution for rapid deployment,installation,update,and deletion of services on edge nodes,and edge container has gradually become the de facto standard for deployment of applications in edge environments.However,due to the resource shortage in the edge environment,the container still has the following problems to run faster and safer at the edge.First of all,the corresponding image is required to deploy the service container.The speed of obtaining the image directly affects the speed of building the service container.However,in the edge environment,the network resources are insufficient and it is difficult to download images from the cloud.The limited storage space of edge devices cannot support caching all images to this node.This leads to a lot of delays when edge nodes build containers,which affects the quality of service.In addition,edge nodes are vulnerable to attack,intrusion,hijacking,or affected by the external environment,resulting in the data consistency of distributed applications deployed by the container being compromised.In view of the above problems,based on the existing research results,this paper carried out in-depth research on the fast supply and reliable guarantee technology of containers in edge computing,and its main work and innovation are as follows:1.This paper proposes a container acceleration supply technology for the edge environment,aiming at the scenario where the image download speed is slow due to limited network bandwidth and low node storage resources.Firstly,this paper designs an image accelerated download scheme DCID based on distributed cache,which integrates the storage resources of all nodes in the edge cluster by means of distributed cache,and designs and implements the image cache and cache acquisition mechanism with the granularity of image layer.Secondly,this paper proposes a cache placement algorithm based on primal duality decomposition to solve the problem of cache placement in image layer files in an edge environment,and further improve the download speed of images through the reasonable allocation of the cache location.Experimental results show that this scheme can significantly accelerate image download speed in an edge environment.Compared with the traditional distributed cache scheme,the DICD image download scheme makes full use of the reusability of image layers,and its acceleration effect and cache hit ratio is better than the traditional distributed cache scheme.2.In the edge environment,edge nodes are easy to be hijacked by attacks and Byzantine errors are easy to occur,thus destroying the consistency of distributed application data.This paper proposes a byzantine fault-tolerant algorithm Edge-Raft.This algorithm is a redesign of the existing classic raft algorithm for potential byzantine errors in edge environments.With the introduction of digital signature,synchronous log detection,polling elections,inert vote,log three phases synchronous mechanism,make it have the byzantine fault tolerance features at the same time,limit the complexity of the messaging to line level,ensure the edge less than a third of the total number of cluster nodes Byzantine error when still can offer effective services for users.The experimental results show that compared with the existing raft algorithm,the proposed algorithm retains the comprehensibility of the raft algorithm while ensuring the usability and activity of the proposed algorithm in the edge environment.Compared with the existing practical byzantine fault-tolerant algorithms,the proposed algorithm limits the time complexity of message passing to the linear level,which ensures the scalability of the proposed algorithm in a multi-node edge environment.