Research On Network Scanning Scheduling Model Based On Docker

With the occurrence of cybersecurity incidents in recent years,the recent "ransomware" incident once again reminds people of the importance of cybersecurity.Network scanning is an important network security technology,it is a proactive preventive measures that can effectively prevent hacker attacks.Improving the performance of network scanning is an important part of the field of Internet security,the emergence of Docker provides a good platform for network scanning.Docker has developed rapidly in recent years,because it has not only good resource isolation capabilities but also low overhead of virtualization,which has many advantages over traditional virtualization technologies.This article deploys network scanning to Docker clusters to improve the performance of network scanning.When there is a scan task,Docker Swarm will first select the node to create the container,and then select the container for task assignment.This article studies from the following two aspects:On the one hand,Docker Swarm's container scheduling algorithm has serious resource fragmentation issues when assigning containers.To solve this problem,this paper uses the improved differential evolution algorithm as the container scheduling algorithm of Docker Swarm.This algorithm guides the optimization search through the group intelligence generated by the mutual cooperation and competition among the individuals in the group.In this paper,memory,CPU and network throughput rate are taken as the influencing factors of the differential evolution algorithm,and an optimal node can be selected to create a container with full consideration of the resource of each node.Therefore,the resource fragmentation of the node is greatly reduced.At the same time,this paper adds the operator with adaptive mutation ability to the differential evolution algorithm,which effectively solves the premature problem that the differential evolution algorithm has when computing the large-scale data.Finally,through the comparison experiments,it is verified that the algorithm can make full use of the resources of the nodes and reduce the resource fragmentation of the nodes.On the other hand,Docker Swarm's task scheduling algorithm results in unbalanced load on each node,"starving" or overloading a node,and unbalanced node load directly affects network scanning performance.Through the analysis of the traditional load balancing algorithm,it is found that the traditional load balancing algorithm is not suitable as the task scheduling algorithm of Docker Swarm.The ant colony optimization algorithm is outstanding in cloud computing load balancing.Therefore,this paper proposes to apply ant colony optimization algorithm to the scheduling model of Docker-based network scanning.The algorithm defines the calculation formula of each node resource,and adopts the pheromone updated formula of the elitist ant colony system to solve the problem that the ant colony optimization algorithm can not solve the large-scale problem when it converges slowly.At the end of this paper,the scanning time and the resource load of each node are taken as performance indexes.The contrast experiments verify that the network scanning task scheduling strategy based on ant colony optimization algorithm makes the load of each node more balanced.