Intelligent eviction strategy for efficient flow table management in OpenFlow Switches

Challa, Rajesh; Lee, Yongseung; Choo, Hyunseung

doi:10.1109/netsoft.2016.7502427

Cited by 24 publications

(22 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Both types are used to drive the parameter tuning in our emulations to decide on the proper actions needed at each state in order to maximize the long-term accumulated rewards. Our emulation results show an improvement of around 60% in reducing the long-term control overhead, and around 14% increase in the table-hit ratio compared to the Multiple Bloom Filters (MBF) method [9] given a fixed size flow table of 4KB.…”

Section: Introductionmentioning

confidence: 93%

SDN Flow Entry Management Using Reinforcement Learning

Al-Fuqaha

Shuaib

et al. 2018

ACM Trans. Auton. Adapt. Syst.

View full text Add to dashboard Cite

Modern information technology services largely depend on cloud infrastructures to provide their services. These cloud infrastructures are built on top of Datacenter Networks (DCNs) constructed with high-speed links, fast switching gear, and redundancy to offer better flexibility and resiliency. In this environment, network traffic includes long-lived (elephant) and short-lived (mice) flows with partitioned/aggregated traffic patterns. Although SDN-based approaches can efficiently allocate networking resources for such flows, the overhead due to network reconfiguration can be significant. With limited capacity of Ternary Content-Addressable Memory (TCAM) deployed in an OpenFlow enabled switch, it is crucial to determine which forwarding rules should remain in the flow table and which rules should be processed by the SDN controller in case of a table-miss on the SDN switch. This is needed in order to obtain the flow entries that satisfy the goal of reducing the long-term control plane overhead introduced between the controller and the switches. To achieve this goal, we propose a machine learning technique that utilizes two variations of Reinforcement Learning (RL) algorithms—the first of which is a traditional RL-based algorithm, while the other is deep reinforcement learning-based. Emulation results using the RL algorithm show around 60% improvement in reducing the long-term control plane overhead and around 14% improvement in the table-hit ratio compared to the Multiple Bloom Filters (MBF) method, given a fixed size flow table of 4KB.

show abstract

Section: Introductionmentioning

confidence: 93%

SDN Flow Entry Management Using Reinforcement Learning

Al-Fuqaha

Shuaib

et al. 2018

ACM Trans. Auton. Adapt. Syst.

View full text Add to dashboard Cite

show abstract

“…Similarly, exchanges of messages between switches lack the authentication/encryption mechanism in the SDN data plane [44]. Toward addressing the aforementioned security challenges, efforts were made lately to address several SDN security attacks [46][47][48].…”

Section: Sdn Security Threat and Vulnerabilitiesmentioning

confidence: 99%

“…This is necessary to avoid removing an entry that may be needed in the near future. Alternatively, instead of relying entirely on the flow expiry period, Challa et al [48] proposed to efficiently manage the flow entries in the flow table through intelligent eviction mechanism based on data logging using multiple bloom filters (MBF) data structure to decide flow entry removal. Although the approach made a step further to remove the unused entries.…”

Section: Timeout and Eviction Mechanismsmentioning

confidence: 99%

Software Defined Networking Flow Table Management of OpenFlow Switches Performance and Security Challenges: A Survey

et al. 2020

View full text Add to dashboard Cite

Software defined networking (SDN) is an emerging network paradigm that decouples the control plane from the data plane. The data plane is composed of forwarding elements called switches and the control plane is composed of controllers. SDN is gaining popularity from industry and academics due to its advantages such as centralized, flexible, and programmable network management. The increasing number of traffics due to the proliferation of the Internet of Thing (IoT) devices may result in two problems: (1) increased processing load of the controller, and (2) insufficient space in the switches’ flow table to accommodate the flow entries. These problems may cause undesired network behavior and unstable network performance, especially in large-scale networks. Many solutions have been proposed to improve the management of the flow table, reducing controller processing load, and mitigating security threats and vulnerabilities on the controllers and switches. This paper provides comprehensive surveys of existing schemes to ensure SDN meets the quality of service (QoS) demands of various applications and cloud services. Finally, potential future research directions are identified and discussed such as management of flow table using machine learning.

show abstract

“…Practically, the size of TCAM-based flow-table in an SDN switch is restricted to a few hundred or at most thousand [11] [12] [13] of entries due to manufacturing cost and high-power consumption. The limited size of the TCAM-based flowtable is not sufficient to manage the massive data flows in a large-scale network, which may result in the flow-table overflow problem [9] [11], and decreases the feasibility to implement a large-scale SDN [14] [15]. But the SDN controller might need to install at least one flow-entry per flow in each switch along the end-to-end path used by the flow to realize per-flow based fine-grained traffic control.…”

Section: Introductionmentioning

confidence: 99%

“…But the SDN controller might need to install at least one flow-entry per flow in each switch along the end-to-end path used by the flow to realize per-flow based fine-grained traffic control. This mechanism will consume the available TCAM space on SDN switches quickly under heavy traffic load conditions [15] [16]. Therefore, as the network scale increases, the available TCAM space will be consumed quickly by all the flow-entries generated during the operation of the SDN network, which will further restrict the network scale and application diversity.…”

Section: Introductionmentioning

confidence: 99%

Deploying a Fast Detection and Eviction Mechanism of Invalid Connection-Oriented Flow-Entries in SDNs: A Scalability Approach

2020

View full text Add to dashboard Cite

As a new paradigm of network architecture, Software Defined Networking (SDN) has been used in a large number of scenarios because it realizes flexible and efficient fine-grained flow control of the network, and promotes the evolution of the network to a programmable and scalable direction. However, the transition of the traditional networking model to SDN architectures poses scalability issues due to the limitation of the flow-table in size. Facing the traffic explosion on future networks with resourceconstrained architectures, the storage space of the flow-table is not enough to bear so many flow-entries so that it not only causes performance degradation in data delivery but also results in scalability and costefficiency issues. To address this issue, in this article, we propose a solution to expedited evict the invalid flow-entries by detecting the disconnect messages of connection-oriented protocols such as Transmission Control Protocol (TCP) and Stream Control Transmission Protocol (SCTP) based on SDN controller and OpenFlow or programming protocol-independent packet processors (P4) switches. The behavior of detection is achieved by adding a specific SDN ruleset within the transport-layer in between the controller and switches. Different from the original timeout solutions, our scheme can delete invalid flow-entries in time according to the transmission layer disconnection instead of relying on the original timeout mechanism. Through a series of simulation results. we also demonstrate the superiority of our proposed solution in reducing the flow-entries occupancy and control overhead on controller, and improving the table-miss rate.

show abstract

Intelligent eviction strategy for efficient flow table management in OpenFlow Switches

Cited by 24 publications

References 10 publications

SDN Flow Entry Management Using Reinforcement Learning

SDN Flow Entry Management Using Reinforcement Learning

Software Defined Networking Flow Table Management of OpenFlow Switches Performance and Security Challenges: A Survey

Deploying a Fast Detection and Eviction Mechanism of Invalid Connection-Oriented Flow-Entries in SDNs: A Scalability Approach

Contact Info

Product

Resources

About