OpenQFlow: Scalable OpenFlow with Flow-Based QoS

Proceedings of the Asia-Pacific Advanced Network, 2011

This paper shows our design of three enhancements to the current OpenFlow technology. OpenFlow is a promising future internet enabling technology that has a great potential to improve the current Internet by providing new functionalities and a new control scheme and thus enabling new smarter applications to be created. Our study aims to provide OpenFlow with three new features; network equipment to equipment flow installation, low level header description, and a new type of inactive flows. Through which we aim to extend OpenFlow's usability, by making it more self-aware and traffic-aware, by relieving some of the load off the OpenFlow controller, by enabling it to have the ability to forward and manipulate user defined protocol headers, and by providing OpenFlow with a method to support flows with strict timing requirements. Those modifications are proposed as a step forward towards encouraging a wider adoption of OpenFlow as an easily programed flowbased network technology that holds a great potential as a future Internet technology that can support newer and smarter class of applications.

2017 Computing Conference, 2017

The increasing demand for network services and quality across wide selections of digital applications in the internet era has caused growing congestion and raised questions about how to deal with prioritizing data in ways tailored to particular uses of applications and managing peak congestion times. Software Defined Network (SDN) in particular Slicing Strategy, seems the best solution due to its new constitution intelligently implemented through the SDN OpenFlow protocol. However, Slicing Strategies specifically "FlowVisor" are limited in certain mechanisms such as Traffic Engineering (TE), which make it a requirement to find new ways to deliver Quality of Service (QoS) for different applications. In this paper, QoSVisor presented as an SDN extension action QoS Slicer based as an enhancement to the standard FlowVisor operation slicing tools to ensure the QoS for each Slice-based class of application.

2013 19th IEEE International Conference on Networks (ICON), 2013

Providing QoS guarantees is an important task in Cloud data centers. Applications running on data centers can achieve predictable performance only if the network resourcebandwidth -is also guaranteed along with the computational and storage resources. While QoS guaranteeing solutions in network has been a subject of intense research in the past decades, in this paper, we focus on studying the capability of software-defined networking architecture in providing QoS guarantees in data centers. In particular, we experiment with the most well-known and standardized OpenFlow, to analyze the goodness of the QoS APIs supported in the recent version (OpenFlow v1.3) in providing rate-guarantees to TCP flows in a data center. We experiment with the commonly implemented TCP congestion control algorithms, namely CUBIC and New Reno. Our study demonstrates that, with rate-limiters, the TCP flows that go through an OpenFlow-enabled switch can experience drops to batches of packets, forcing them to switch to the slowstart phase periodically, thereby leading to inefficient bandwidth utilization. This behaviour can adversely affect the performance of applications generating the flows. Hence, we need to develop better mechanisms that limit flow-rates while allowing the flows to operate in their congestion-avoidance phase.

Proceedings of the Asia-Pacific Advanced Network, 2013

This paper shows our proposed design and discusses in more details about the evaluation of our two enhancements to the current OpenFlow technology [11]. OpenFlow is a promising future internet enabling technology that has a great potential to improve the current Internet by providing new functionalities and a new control scheme, and thus, enabling new smarter applications to be created. Our study aims to provide OpenFlow with two new features; network equipment to equipment flow installation (Ne-Ne FI), and a new type of proactive flows. Through which we aim to extend OpenFlow's usability, by making it more self-aware and traffic-aware, by relieving some of the load off the OpenFlow controller, and by providing OpenFlow controller with a relaxed method to support flows with strict timing requirements.

Cluster Computing, 2019

Datacenter applications (e.g., web search, recommendation systems, and social networking) are designed to have a high fanout for the purpose of achieving scalable performance. Frequent fabric congestion (e.g., due to incast, imperfect hashing) is a corollary of such a design. This is true even when the network utilization is low. Such fabric congestion exhibits both temporal as well as spatial (intra-rack and inter-rack) variations. There exist two basic design paradigms which are used to address this issue. Current solutions lie somewhere between the two. On one hand we have arbiter based approaches where senders poll a centralized arbiter and collectively obey global scheduling decisions. On the other end of the spectrum, we have self adjusting end point based approaches where senders independently adjust transmission rate based on network congestion. The former incurs greater overhead, compared to the latter which trades off complexity for sub-optimality. Our work seeks a middle ground-optimality of arbiter based approaches with the simplicity of self adjusting end point based approaches. Our key design principle is that since the receiver has complete information regarding the flows destined for it, rather than having a centralized arbiter schedule flows or the senders making independent scheduling decisions, the receiver can orchestrate the various flows destined for it. Since multiple receivers may be using a bottleneck link, datapath visibility should be used to ensure fair sharing of the bottleneck capacity between receivers with minimum overhead. We propose RecFlow, which is a receiver-based proactive congestion control scheme. RecFlow employs OpenFlow provided path visibility to track changing bottlenecks on the fly. It spaces TCP acknowledgements to prevent traffic bursts and ensure that no receiver exceeds its fair share of the bottleneck capacity. The goal is to reduce buffer overflows while maintaining fairness among flows and high link utilization. Using extensive simulation results and real testbed evaluation, we show that compared to the state-of-the-art, RecFlow achieves up to 69 improvement in the inter-rack scenario and 1.59 in the intra-rack scenario while sharing the link capacity fairly between all flows.

ICTC 2012

OpenFlow has been envisioned as a promising approach to next-generation programmable and easyto-manage networks. However, the inherent heavy switchcontroller communications in OpenFlow may throttle controller responsiveness and, ultimately, network scalability. In this paper, we identify that a key cause of this problem lies in flow setup, and propose a Control-Message Quenching (CMQ) scheme to address it. CMQ requires minimal changes to OpenFlow, imposes no overhead on the central controller which is often the performance bottleneck, is lightweight and simple to implement. We show, via worst-case analysis and numerical results, an upper bound of performance improvement that CMQ can achieve, and evaluate the average performance via experiments using a widely-adopted prototyping system. Our experimental results demonstrate considerable enhancement of controller responsiveness and network scalability by using CMQ, with reduced flow setup latency and elevated network throughput.

Journal of Networks, 2015

This paper proposes a method for optimizing bandwidth usage in Software Defined Networks (SDNs) based on OpenFlow. Flows of small packets presenting a high overhead, as the ones generated by emerging services, can be identified by the SDN controller, in order to remove header fields that are common to any packet in the flow, only during their way through the SDN. At the same time, several packets can be multiplexed together in the same frame, thus reducing the overall number of frames. The method can be useful for providing QoS while the packets are traversing the SDN. Four kinds of small-packet traffic flows are considered (VoIP, UDP and TCP-based online games, and ACKs from TCP flows). Both IPv4 and IPv6 are studied, and significant bandwidth savings (up to 68 % for IPv4 and 78 % for IPv6) can be obtained for the considered kinds of traffic. The optimization method is also applied to different public Internet traffic traces, and significant reductions in terms of packets per second are achieved. Results show that bandwidth consumption is also reduced, especially in those traces where the percentage of small packets is high. Regarding the effect on QoS, the additional delay can be kept very low (below 1 millisecond) when the throughput is high, but it may become significant for lowthroughput scenarios. Thus, a trade-off between bandwidth saving and additional delay appears in those cases.

2017

This paper provides a proof of concept of an SDN Application to provide QoS for real-time services on SDN networks. Common real-time services are for example VoIP or M2M protocols like OPC UA and MQTT. We provide a proof of concept in the form of a specialized application for SIP traffic. The application lowers the latency and call-setup-time for VoIP calls. The application uses the metering and queueing features of OpenFlow 1.3 to assure high quality of service. The evaluation and optimization of the application is still in progress.

Computer Communications, 2020

To ensure the quality of service of an end-to-end connection, current network solutions are mostly dependable on the differentiation between different classes of traffic. The Software-defined networking (SDN) architecture has emerged to offer network programmability, giving to network operators a programmatic control over their network. In SDN, network devices are programmed in many ways, having a standard, open, and vendoragnostic interface, e.g., OpenFlow, enabling the control plane to instruct the forwarding behavior of network devices from different vendors. In this paper, we introduce the Programmable Labels (ProgLab) approach to support traffic differentiation with QoS guarantees as a low-cost alternative built over an SDN architecture. The idea relies on the simplification of a packet-forwarding operation which relies on the remainder of a division, instead of classical table lookup method. ProgLab computes programmable label at the control plane by solving a congruence system from Residue Number System and the co-prime numbers assigned to the switches in the path of an end-to-end connection. Such label has a meaning within this network that expresses the entire route, addressing the respective traffic class at each switch's logical queue along the path. ProgLab approach has been implemented through the P4 language and evaluated through an emulation-based evaluation. The experiments demonstrated the feasibility of ProgLab and showed its ability in providing QoS differentiation on demand.

2018

This paper provides an overview of the Quality of Service (QoS) capabilities defined in the OpenFlow specification. Several vendor documentations from off-the-shelf products are compared with the OpenFlow specification. This research reveals inconsistencies between the specification and the vendors implementation. Queues for examples are not implemented by all vendors. This gap can lead to interoperability problems in a network while using hardware from different vendors. The research also shows, that the majority of vendors provide a port statistic function which gives information about incoming and outgoing bandwidth about each port of a switch. Based on this function a QoS aware routing application for off-the-shelf switches is proposed. This concept can be used to change the flow of traffic in an OpenFlow network based on the utilization of the interfaces. Based on the conducted research, the application can be used with hardware from multiple vendors. This paper does not contai...