Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-20 Cooperative journals: 《计算机应用研究》
Abstract： In order to solve the problems of poor load balancing caused by dynamic scheduling in data center networks(DCN) , this paper proposed a dynamic load balancing algorithm based on flow scheduling selection(DLBFSS) . Firstly, DLBFSS calculates the equal-cost shortest path for each large flow on the congested link and removes the path with insufficient bandwidth. Then, the available throughput of each remaining path for large flows was calculated, and the path with maximum available throughput was selected as the optimal scheduling path. Finally, the congestion probability of each large flow was defined by considering the flow bandwidth and the load of the optimal path, and the congestion probability was used for large flow scheduling selection. The experimental results show that DLBFSS can reduces the network delay, improves the flow bandwidth utilization and ensures load balancing compared with ECMP(equal-cost multi-path) and typical large flow scheduling algorithms.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-20 Cooperative journals: 《计算机应用研究》
Abstract： To solve the problems of network congestion and load imbalance caused by elephant flow that carries large amounts of data in data center networks. This paper proposed an SDN based load balancing mechanism of elephant flow(EFLB) . When network load exceeded the threshold, the controller split the detected elephant flows down into multiple mice flows by Openflow feature and calculated dynamically the minimum load switch according to the collected network topology and link states to ensure load balancing. Experimental results show that the EFLB achieves network load balancing by improving the network throughput and link utilization compared with ECMP((equal-cost mulit-path routing) .
Hits
YES
Hits
Downloads
Comment