Research And Implementation Of Resource Scheduling Management Technology Of Tactical Cloud Architecture

The rapid development of technologies such as cloud computing and the Internet of Things has caused changes in the way of air combat.The intelligence and unmanned use of weapons have become important research trends in the field of air combat.In response to the need of intelligent systems for collaborative work in future combat systems,various technical researches on tactical cloud architecture have been carried out.In tactical cloud,resource scheduling and management provides the linkage capabilities of the modules internally or between them,and balance the load of resource services,while also monitoring the status changes and execution processes of resource services.Therefore,a resource scheduling and management module needs to be designed,which makes the load of the entire platform more balanced,operates more smoothly,and provides important guarantees for building a tactical cloud platform.However,the platform interface in the tactical cloud architecture has heterogeneous problems.The inconsistent definition of the interface and the different development languages between different modules restrict the data exchange and service calls between the resource scheduling and management module and other modules.Therefore,a cross-language and standard unified interface design rule needs to be formulated.This thesis uses REST style to formulate interface design rules for tactical cloud,and solves the problem of inconsistent platform interface design.This rule details the standards that should be followed in the interface design process from four aspects: classification,readability,ambiguity,and reusability.According to the rules,the interfaces to be used in the resource scheduling and management module are defined,and detailed interface design content is given.For resource scheduling of tactical cloud,this thesis designs a weighted random selection load balancing service and an interface analysis service with a two-layer load balancing architecture.The core of the load balancing service is a weighted random selection algorithm,which provides call interfaces for other modules and solves the load balancing problem of the entire platform.The interface analysis service combines load balancing services and Dubbo’s built-in load balancing algorithms,while integrates two design patterns of client and server to achieve a two-layer load balancing architecture,which solves the load balancing problem of resource calls.At the same time,the "bottleneck" of load balancing is avoided.The test results show that the interface analysis service can balance user requests to different services,and the process has passed the load balancing services.For resource management of tactical cloud,this thesis designs two functional modules of service monitoring and task management.In service monitoring module,a "heartbeat" detection sub-module is designed by the PING command,and a load monitoring sub-module is designed by periodically calculating server performance data,which solves the problem of resource service status detection and load monitoring.In task management,according to the resource characteristics and runtime status information,a secondary resource lock competition mechanism and a listener control mechanism are designed,which ensure the task sequence in the tactical cloud can be executed in a complete and orderly manner.The test results of the service monitoring module show that the module can correctly detect failed nodes and overloaded nodes.The test results show that the service monitoring module can correctly detect failed nodes and overloaded nodes,and the task management module can determine the resource access permissions of the task sequence through the competition mechanism.Then,the sequence of obtaining the permissions completes the corresponding task execution under the control mechanism.