The cloud has become an indispensable infrastructure for modern enterprises, and cloud-based cross-data center (DC) services are becoming more prevalent. However, most of the existing congestion control (CC) algorithms are designed for individual networks and are ill-suited for cross-DC networks for two reasons. First, intra-DC and cross-DC flows have different control loop lengths. Second, they can hardly share the same parameter settings for buffer-based signals. Concentrating on accurate and rapid CC for cross-DC networks, in this paper, we propose Gear, a novel INT-based CC mechanism that assigns different tasks to intra- and cross-DC flows. With the help of In-Band Network Telemetry (INT), intra-DC flows can determine both their appropriate sending rates and the bandwidth occupation of cross-DC flows, which facilitates their ability to fully utilize the network while keeping low latency. Meanwhile, the cross-DC flows are responsible for ensuring fair bandwidth share between intra-DC and cross-DC flows through proactive bandwidth allocation. Our experimental results demonstrate that Gear significantly optimizes the throughput fairness under cross-DC networks compared with state-of-the-art CC algorithms. Gear also improves throughput by up to 20%.

Congestion control schemes based on information, such as queue size and traffic load, have become increasingly important, especially in data center networks, where the applications have stringent bandwidth and latency requirements. However, some congestion control schemes employ In-band Network Telemetry (INT), which introduces nontrivial bandwidth overhead. In this paper, we propose an efficient congestion control scheme: Precise ECN Encoding (PEE). PEE refactors the Explicit Congestion Notification (ECN) marking logic and proposes a novel ECN-based multi-packet joint encoding/decoding mechanism to enable more precise congestion perception. Compared with the state-of-the-art schemes, PEE achieves precise congestion control without introducing extra bandwidth overhead, achieving high throughput and low latency simultaneously. Comprehensive experimental results show that, compared to TIMELY, DCQCN, and HPCC, PEE reduces the average Flow Completion Time (FCT) by 102.1%, 26.8%, and 16.6% respectively under 80% network load.

In recent years, hybrid congestion control (CC) algorithms that combine rule-based CC and learning-based CC have gained significant attention. They incorporate the fast adaption ability of learning-based CC and the stability of rule-based CC, tending to select the better-performing rate based on the network feedback. However, the practical implementation of such algorithms has revealed primary issues. Specifically, they require both CCs to run alternately, which results in a poorly performing CC continuing to run in the network. Moreover, hybrid CCs cannot converge to the optimal rate when both CCs perform poorly. This paper proposes Anole to address these issues. Anole has three main algorithmic contributions: 1) Anole always selects the better-performing CC, 2) Anole temporarily deprecates the consistently underperforming CC, 3) when both CCs pervform poorly, Anole infers the optimal sending rate based on the network feedback. We carry out comprehensive experiments in both emulated and real-world wired networks, as well as in real-world WiFi networks, to assess the performance of Anole. The experiment results demonstrate that Anole achieves approximately 6% higher throughput in real-world links and 34% lower delay in the 48Mbps link compared to the state-of-the-art CC. Anole also exhibits superior performance in adaptability and fair convergence.

Much effort has been devoted to improving the performance of network transmission. Yet, the impact of packet size which is limited by the 1500-byte maximum transmission unit (MTU) has not received adequate attention. Through comprehensive experiments, we find that jumbo frames which are commonly used as an alternate do not always yield the best performance under different transmission situations.In this paper, we elaborate on the limitations of the regular and jumbo frames and analyze how packet sizes affect network performance. Based on these, we present Adaptively Packet Sizing (APS), a dynamic packet size adjustment method that can be easily integrated into existing window-based congestion control algorithms. APS utilizes a machine learning method to predict the optimal packet size, which can minimize flow completion time (FCT) according to the instantaneous network condition. Besides, a packet size based priority mechanism is proposed to further improve the performance. We implement APS in both simulation and testbed environments. APS reduces the FCT by up to 50% and gains better performance in scenarios with various loss rates.

Datacenters currently deploy shallow-buffered switches to achieve low latency by avoiding long waiting time in the data plane. However, the limited buffer space in the switch causes the frequent overflow and the notorious TCP incast problem. Moreover, the simple scheduling strategy in buffer deprives the switch of the ability to offer deeply differentiated services. Therefore, we present a novel priority-based flow-aware in-network caching scheme, named Poche, which supplies more control capabilities for the network side through introducing some additional cache resource into switches. Poche classifies network traffic into multiple priorities according to the latency requirements of flows. The end server adds priority tags to packets and sets different RTO values for flows with distinct priorities. The switch monitors the buffer utilization of each port and performs the priority-based flow-aware caching and injecting strategies based on the analysis of the scheduling model between the buffer and cache. We conduct comprehensive experiments to compare Poche with the state-of-the-art traffic optimization schemes. The results demonstrate that Poche can reduce the FCTs of latency-sensitive flows by at least 59.1% and improve the network throughput by at least 54.4%, while ensuring the finite cached volume and effectively addressing the incast problem.

Many cloud applications in modern datacenters have very demanding latency requirements, making flow completion time (FCT) an important metric for evaluating the network performance. Existing network flow scheduling methods either base on pre-known information or have poor performance. Therefore, we present LAFS, an efficient learning-based flow scheduling approach which minimizes the FCT with estimated information of flows. LAFS combines system call monitoring and learning methods to learn the flow size and implements the Shortest Remaining Processing Time (SRPT) principle with in-network priorities. Moreover, LAFS adopts flowlets to alleviate the packets disorder problem in fine-grained flow scheduling. Our theoretical analysis and extensive simulations show that LAFS is a practical design and significantly outperforms other information-agnostic designs like DCTCP and PIAS under diverse workloads.