Efficient load-balancing mechanisms are critical for maximizing performance and increasing the quality of service (QoS) of data center networks (DCNs). Obtaining the optimal QoS while minimizing resource consumption remains a significant challenge. This paper proposes the streamlined pathway (SP) model, which is a flow scheduling solution that requires minimal statistical knowledge of the DCN data plane. The SP model utilizes the software-defined networks (SDN) paradigm with less information gathered from the DCN data plane, besides the traditional hash-based flow scheduling mechanism, the Equal Cost Multi-Path (ECMP). In SDN, the proposed methodology harnesses a minimal yet powerful set of statistical data extracted from the DCN data plane, including port throughput and elephant flow information on the aggregate switches of the DCN fat-tree topology. Several experiments, in addition to theoretical analysis, have been conducted to demonstrate the efficiency of the proposed SP model in terms of QoS enhancement. These results confirm that SP outperforms leading techniques such as Sieve, Hedera, and ECMP, concerning bisection bandwidth, DCN link utilization, packet loss, and packet delivery latency.

翻译：高效的负载均衡机制对于最大化数据中心网络（DCN）的性能与提升服务质量（QoS）至关重要。在最小化资源消耗的同时获得最优的QoS仍是一项重大挑战。本文提出了流线化路径（SP）模型，这是一种仅需DCN数据平面最少统计知识的流调度解决方案。该SP模型利用软件定义网络（SDN）范式，除了传统的基于哈希的流调度机制——等价多路径（ECMP）之外，它从DCN数据平面收集的信息更少。在SDN中，所提出的方法利用了从DCN数据平面提取的一组最小但功能强大的统计数据，包括DCN胖树拓扑中汇聚交换机的端口吞吐量和大象流信息。除了理论分析外，还进行了多项实验，以证明所提出的SP模型在提升QoS方面的效率。这些结果证实，在二分带宽、DCN链路利用率、丢包率和数据包交付延迟方面，SP模型优于Sieve、Hedera和ECMP等领先技术。