Font Size: a A A

Research On Load Balancing Technology Of Service Integration Framework

Posted on:2020-11-13Degree:MasterType:Thesis
Country:ChinaCandidate:W QuFull Text:PDF
GTID:2428330623459909Subject:Computer technology
Abstract/Summary:PDF Full Text Request
The Service Integration Framework is a DDS-based SOA implementation solution that provides a platform for application resource sharing and management for distributed realtime application systems,and solves the problem of communication performance.However,due to the fact that there are not multiple copy management and load balancing mechanisms in the service integration framework,it is unable to cope with the application scenarios of open and large-scale service requests.Aiming at the above problems,the paper proposes a load balancing solution based on service dynamic deployment under the service integration framework,which shortens the service response delay and improves the utilization of system resources.The main contributions of the paper are:1.The load balancing architecture in the service integration framework is proposed to support multi-copy dynamic deployment and load balancing of services.The service state management model is extended,while at the same time the service state is defined,and the transition process between service states is agreed on.A standard service state management interface is provided in the service integration framework to implement service state transitions.2.key mechanisms and strategies are put forward to support dynamic deployment and load balancing of services.The service status management service and node load metrics are provided;The service replica state change implementation mechanism based on DDS interceptor is designed.In order to make the service in the system meet the service level agreement(SLA)constraints and avoid the waste of system resources,a service copy number determination algorithm based on the service request rate constraint is designed;In order to ensure the stable deployment of services and improve the resource utilization in the system,a target node selection algorithm based on node load prediction and resource constraints is designed.A service reference maintenance method transparent to users is proposed for the problem of service reference invalidation after dynamic deployment of services.Aiming at solving the service state consistency maintenance problem in the service migration process,a general service state data transfer mechanism based on byte stream is proposed,so that different state data can be transferred by the same mechanism.3.Based on the aforementioned load balancing solution,a load balancing prototype system in the Service Integration Framework is designed and implemented,and the prototype system is tested for function and performance.The experimental results show that the load balancing solution for the Service Integration Framework proposed in the thesis can better cope with the application scenarios of large-scale and dynamic changes in concurrent service requests in large-scale open distributed systems,shorten service response time and optimize system resources utilities.
Keywords/Search Tags:Service Integration Framework, Load Balancing Architecture, Service State Management Model, Service Dynamic Deployment
PDF Full Text Request
Related items