随着云计算的大规模普及，负载均衡已经以用户服务的形式出现在云平台上，但传统的负载均衡算法不能利用云平台动态管理计算资源的优势。基于预测的动态负载均衡算法根据当前云平台应用负载特征使用BP神经网络对负载进行实时预测，然后根据预测的负载变化趋势提出负载均衡器后端实例调整策略。实验结果表明，基于预测的动态负载均衡算法能够在满足用户SLA的前提下，促进资源节约。%With the massive popularity of cloud computing ,load balancing has appeared in the form of customer service on cloud platform .But the traditional load balancing algorithm can not take advantage of the ability of dynamically manage re‐sources on cloud computing platform .Dynamic prediction load balancing algorithm based on BP neural network which based on the current application load for real‐time load forecasting ,then made back‐end load balancer instance adjust poli‐cy based on trends of load that predicted .Experimental results show that the prediction of dynamic load balancing algo‐rithm can satisfy the user SLA ,saving resources .
展开▼
