Cloud calculating load balancing scheduling algorithm based on double-weighted least-connection algorithm

Abstract

The invention discloses a cloud calculating load balancing scheduling algorithm based on a double-weighted least-connection algorithm and belongs to the network technology field. According to the cloud calculating load balancing scheduling algorithm, after weighting of performances of resource servers through the weighted least-connection scheduling algorithm, real-time information such as CPU idleness and memory idleness on the servers is utilized to dynamically express weights of the performances of the servers, so residual processing capability of each node server can be fully assessed and utilized; corresponding weights are given to tasks according to complexity degrees of task types; no relatively large inclination of a load of each node is guaranteed when a system is in a long time operation state, summation of all task weights of each server and weight ratio of performances of the servers are calculated by a scheduler before task distribution, new tasks are distributed to the servers with smallest ratios. The cloud calculating load balancing scheduling algorithm shortens average accomplishment time of the cloud calculating service system, improves system efficiency and further improves a load balancing degree of each resource server in a cloud data center.

Description

technical field [0001] The invention belongs to the field of network technology, and more precisely, the invention relates to a cloud computing data center load balance scheduling strategy method. Background technique [0002] The cloud computing data center has hundreds of millions of servers or PCs, etc., with a large number of resources and strong heterogeneity. Among them, the load balancing problem of the data center server directly affects the performance of the whole system, so the research on the load balancing scheduling algorithm has become a hot spot of people's attention and research. [0003] The cloud computing data center dynamically allocates resource servers to users according to user needs and scheduling policies. Due to the inconsistency in data center server resource processing performance or inconsistency in user demand specifications (for example, the length of time for users to request services, the difficulty of problems), if a simple allocation sche...

AI Technical Summary

Problems solved by technology

This algorithm is suitable for situations where the processing performance of server nodes is different, but there are still several shortcomings: (1) The setting of server performance weights cannot accurately reflect the real processing capabilities of nodes

Since the weight value is set in advance according to the configuration of the server nodes and the experience of the administrator, as the running time increases, some nodes are already heavily loaded, but they still receive connection requests continuously, while some nodes are idle. As a result, these pre-set weights can no longer accurately reflect the real processing capabilities of the nodes, causing the load of the entire system to tilt, affecting the performance of the system

(2) Only the number of connections is used to represent the node load, which cannot accurately represent the real-time load of the service node

Therefore, it is not comprehensive to measure the load only by the number of connections

Images