CAJ | 학술논문

传统基于应用层中间件或代理的服务迁移技术容易造成单点故障且扩展性差，并且流媒体的交互性和长会话特性使其难以实现对用户透明。为此，提出一种基于OpenFlow的流媒体云服务迁移方法。通过在网络层进行流媒体云迁移实现对用户的服务透明，同时针对网络层的服务迁移代价问题，提出一种基于用户请求预测的迁移策略，在提升系统容量的同时优化服务迁移数。实验结果表明，该迁移策略可有效地提高服务请求接受率，进而增强系统服务能力。
전통기우응용층중간건혹대리적복무천이기술용역조성단점고장차확전성차，병차류매체적교호성화장회화특성사기난이실현대용호투명。위차，제출일충기우OpenFlow적류매체운복무천이방법。통과재망락층진행류매체운천이실현대용호적복무투명，동시침대망락층적복무천이대개문제，제출일충기우용호청구예측적천이책략，재제승계통용량적동시우화복무천이수。실험결과표명，해천이책략가유효지제고복무청구접수솔，진이증강계통복무능력。
Traditional migration techniques based on a middleware or a proxy at the application layer are likely to cause a single point of failure and have poor scalability,and it is difficult for them to be transparent to users because of the features of streaming media interaction and long session. Aiming at the problems,this paper proposes an OpenFlow-based service migration method for streaming media cloud. The method implements service migration at the network layer which is transparent to users. In the meantime,in view of the cost of service migration at the network layer,this paper proposes a request prediction-based migration strategy which can improve the system capacity while optimize the quantities of service migration. Experimental results show that the strategy can effectively increase the acceptance rate of requests,thus improve the service capability of the system.