A port-based forwarding load-balancing scheduling approach for cloud datacenter networks

Abstract Today’s datacenter networks (DCNs) scale is rapidly increasing because of the wide deployment cloud services and rapid rise edge computing. The bandwidth consumption cost a DCN are growing sharply with extensions network size. Thus, how to keep traffic balanced key challenging issue. However, traditional load balancing algorithms such as Equal-Cost Multi-Path routing (ECMP) not suitable for high dynamic in DCNs. In this paper, we propose port-based forwarding scheduling (PFLBS) approach Fat-tree based DCNs some new features which can overcome disadvantages existing methods following aspects. Firstly, define source-routing addressing scheme, decreases switch complexity makes table-lookup operation unnecessary. Secondly, on proposed an effective mechanism obtain multiple available paths flow Fat-tree. All path information saved servers each server only needs maintain its own information. Thirdly, efficient algorithm implement large flows dynamically terms current link utilization ratio. This method computing, reduce switches power whole network. experiment results indicate that PFLBS has better performance compared ECMP, Hedera MPTCP approaches, completion time improves average throughput significantly. simple be implemented few signaling overheads.