HPSTOS: High-Performance and Scalable Traffic Optimization Strategy for Mixed Flows in Data Center Networks

Abstract

In data center networks, traffic needs to be distributed among different paths using traffic optimization strategies for mixed flows. Most of the existing strategies consider either distributed or centralized mechanisms to optimize the latency of mice flows or the throughput of elephant flows. However, low network performance and scalability issues are intrinsic limitations of both strategies. In addition, the current elephant flow detection methods are inefficient. In this paper, we propose a high-performance and scalable traffic optimization strategy (HPSTOS) based on a hybrid approach that leverages the advantages of both centralized and distributed mechanisms. HPSTOS improves the efficiency of elephant flow detection through sampling and flow-table identification. HPSTOS guarantees preferential transmission of mice flows using priority scheduling and adjusts their transmission rate by coding-based congestion control on the end-host, reducing their latency. Additionally, HPSTOS schedules elephant flows by cost-aware dynamic flow scheduling on a centralized controller to improve their throughput. The controller handles only elephant flows, which constitutes the minority of the flows, allowing effective scalability. Evaluations show that HPSTOS outperforms existing schemes by realizing efficient elephant flow detection and improving network performance and scalability.

Traffic‐aware rate control for mix‐flow in datacenter

“…The distributions of flow sizes follow the heavy-tailed characteristics, with over 90% of the total bytes being contributed by less than 10% of large flows. These flows are randomly transmitted between selected source and destination hosts, and their arrival times follow a Poisson process [13,16]. By default, we set the flows as deadline flow or non-deadline flow randomly.…”

“…Baselines and parameter : We compare TRC, DCTCP [6], the strict priority queue (SPQ) equipped with eight queues, and the state‐of‐the‐art deadline‐aware schemes, including pFabric [11], Karuna [13], DA&FD [14], and HPSTOS [16]. Note that within the SPQ, the transmission protocol is DCTCP, and the high‐priority queues, ranging from 0 to 6, follow an Earliest Deadline First (EDF) strategy to serve deadline flows.…”

“…To improve the above mix-flows' performance, it is important to design a new rate control method to reduce the deadline missing ratio for deadline flows, while minimizing the average FCT for non-deadline flows. To meet the requirements of deadline flows, the researchers have proposed a series of deadline-aware schemes, including D 2 TCP [9], PDQ [10], pFabric [11], D 3 [12], Karuna [13], DA&FD [14], mixCoflow [15], and HPSTOS [16]. These schemes can be categorized into two main categories: best-effort schemes and near-deadline schemes.…”

Integrating Resilient Tier N+1 Networks with Distributed Non-Recursive Cloud Model for Cyber-Physical Applications

“…The work employed multiple transceivers (TRXs) with a wavelength selective switch (WSS) reprogrammed with an SDN control plane. In [31], the authors developed a DCN that leverages a high-performance and scalable traffic optimization strategy (HPSTOS). This depends on a hybrid scheme that leverages the benefits of both centralized and distributed mechanisms to enhance the efficiency of flow detection via sampling and flow-table recognition.…”

FMap: A fuzzy map for scheduling elephant flows through jumping traveling salesman problem variant toward software‐defined networking‐based data center networks

“…Therefore, DCNs demand an improved network performance and availability. [1][2][3][4] Moreover, to achieve high performance in DCNs, some applications leverage the pattern of many-to-one communication. However, the high concurrency leads to problems such as transmission control protocol (TCP) incast which in turn might lead to application performance degradation.…”