QPS-ToR boosts datacenter network speed and flow times by improving scheduling
QPS-ToR: A Parallel Iterative Switching Algorithm for Reconfigurable Optical Datacenter Switching
Networking and Internet Architecture
Summary
Data centers use networks to move lots of data between many computers, but slow scheduling systems can limit how fast this happens. The authors improved an existing scheduling method in optical data center networks by using a new algorithm called SW-QPS, which allows faster and smarter data traffic management. Their approach increased data throughput by up to 36% and made data flows finish up to 82% faster in simulated tests. This means data centers can handle more data quickly and efficiently with their method.
What this means in practice
- •For datacenter network operators: Improve network scheduling to increase throughput and reduce data flow delays in optical data center networks using QPS-ToR.
- •For network hardware designers: Design ToR switches that implement sliding-window scheduling to achieve higher throughput and lower latency in optical data center topologies.
Authors
Dongzhao Song, Qianru Yu, Jun Xu
Abstract
Reconfigurable optical data center networks (RODCNs) have emerged as a promising solution for scaling DCN capacity, yet their scheduling mechanisms remain a performance bottleneck: traffic-oblivious schemes inherently limit throughput, while the state-of-the-art traffic-aware scheme, NegotiaToR, uses single-iteration iSLIP as its scheduling engine, which limits throughput to around 60% and treats all source-destination pairs with equal priority regardless of queue length. We propose QPS-ToR, which replaces NegotiaToR's scheduling logic with SW-QPS, a sliding-window algorithm originally proposed for crossbar scheduling that achieves around 90% throughput with a single low-complexity iteration. QPS-ToR operates within NegotiaToR's existing workflow, requiring only a revision of the scheduling cycle from three-step Request-Grant-Accept (RGA) to two-step Request-Grant (RG) with a sliding window mechanism. In flow-level simulations with 128 ToRs under realistic datacenter workloads, QPS-ToR achieves up to 36% higher throughput and 82% lower flow completion time (FCT) compared to NegotiaToR, and consistently outperforms RotorNet, a representative traffic-oblivious scheme, on the parallel network topology.