Method and system for controlling adaptive load-balancing of cloud computing server

Abstract

The invention relates to the field of computers and discloses a method for controlling adaptive load-balancing of a cloud computing server. The method comprises the steps as follows: detecting response time of a service request, calculating the priority of the service request, adding the service request into a service request queue, and dispatching and distributing the service request to a corresponding server; detecting and feeding back node parameter information of the server, calculating node distribution weight of the server, and updating the weight of the server node; dynamically collecting real-time load conditions of each server, analyzing and re-calculating the performance of each server in a cluster, evaluating and monitoring whether there is a network congestion risk or not, and re-distributing the service request in the request queue to the server according to an investigating result. The invention further discloses a system for controlling adaptive load-balancing of the cloud computing server. The method and the system for controlling adaptive load-balancing of the cloud computing server of the invention could balance the load of each application server to the largest extent so as to effectively configure system resources.

Description

technical field [0001] The invention relates to the field of computers, in particular to a cloud computing server self-adaptive load balancing control method and control system. Background technique [0002] In the next generation network, the softswitch technology is adopted to realize the separation of call control and service provision. The application server is introduced into the softswitch system to provide a development and operation platform for third-party services. In order to meet the large-capacity requirements of the business system, multiple application servers are required to form a cluster system. Each business server in the cluster will install a configuration client software, and regularly report its working status to the server called the configuration center. When the configuration center server does not receive a status report from a certain business server within a period of time, It is considered that the business server is faulty, and the business s...

AI Technical Summary

Problems solved by technology

[0005] From the perspective of the access server, the service quality of the service server depends not only on the processing capability of the service server itself, but also on the network bandwidth between the two, network delay, and changes in the external environment of the service server. The existing load balancing algorithm only distributes request packets in proportion to the maximum quota value of the business server, that is, only considers the processing capacity of the business server itself, and when the network leading to the business server is congested, packet loss occurs, or the business server When its own performance degrades, if the access server still distributes request packets according to the ratio of the maximum quota value, this will cause load balancing between business servers cannot be truly achieved

[0006] In the process of realizing the above load distribution, the inventor found at least the following problems in the prior art: the load is evenly distributed to each application server without considering the load capacity and actual load status of each server, which may cause some servers to be long Time is in a high load state, and some servers are idle for a long time, wasting system resources

Images