Preferencje help
Widoczny [Schowaj] Abstrakt
Liczba wyników

Znaleziono wyników: 15

Liczba wyników na stronie
first rewind previous Strona / 1 next fast forward last
Wyniki wyszukiwania
Wyszukiwano:
w słowach kluczowych:  OpenFlow
help Sortuj według:

help Ogranicz wyniki do:
first rewind previous Strona / 1 next fast forward last
1
Content available remote Proposed method for partial node replacement by software defined network
EN
Since it is impractical to replace the entire traditional network by the SDN network due to some constraints i.e. financial budget, limited skills to SDN, in addition to the need to have the benefits and flexibility of the traditional network, the partial replacement implemented by deploying or replacing some legacy nodes by the SDN switches have emerged. Such replacement requires routing and security addressing coordination issues. In this research, we present our proposed solution for automatic replacement of a segment of the legacy network by SDN nodes, and generation of a set of OpenFlow rules and switches configuration that meets the traditional network behavior requirements. The rules are identified based on the analysis of the network traffic acquired from the legacy segment.
PL
Technologia Wieloprotokołowej Komutacji Etykietowej (MPLS – MultiProtocol Label Switching) dostarcza odpowiednie mechanizmy wspierające proces zachowania jakości usług, jednak wymaga nadrzędnego elementu sterującego. Takim elementem może być kontroler zrealizowany w koncepcji Sieci Definiowanych Programowo (SDN – Software Defined Networking), wprowadzający abstrakcję, rozdzielającą płaszczyznę przekazywania od płaszczyzny sterowania. W niniejszym artykule przedstawiono sposób połączenia obu technologii w ramach jednej sieci wielodomenowej.
EN
Multiprotocol Label Switching (MPLS) provides appropriate mechanisms to support the process of maintaining the quality of services. However, it requires a superior control element. Such an element may be a controller implemented according to Software Defined Networking concept, introducing abstraction that separates the forwarding plane from the control plane. This article presents the method of combining both technologies within one multi-domain network.
PL
Artykuł opisuje nietypowe zastosowanie symulatora zdarzeń dyskretnych OMNeT++ do sterowania programowalnymi sieciami komputerowymi (SDN). W środowisku OMNeT++ przygotowano sterownik sieci programowalnych obsługujący protokół Open Flow w wersji 1.0.0. Sterownik może wykorzystywać rzeczywisty interfejs sieciowy. Ponadto sterownik wyposażono w odpowiednie liczniki wiadomości i ich rozmiarów. Napisano także aplikację wspierającą zarządzanie ARP w sieci SDN. Przygotowano środowisko do przeprowadzenia badań, oparte o emulator Mininet i bazę danych MySQL oraz omówiono zebrane wyniki.
EN
The article describes unusual usage of discret event simulator OMNeT++ to control software defined networks (SND). An SDN Controller that supports OpenFlow version 1.0.0 has been prepared in OMNeT++. The Controller may use a real network’s interface. Moreover, the Controller has been equipped in counters of messages and their sizes. An application that supports maintaining of ARP has been prepared. The test environment based on Mininet and MySQL instances has been created. The results have been presented and discussed.
PL
W artykule dokonano analizy procesu zbierania statystyk ze switch’y SDN. Gromadzone pomiary za pomocą protokołu OpenFlow mogą być wykorzystywane przez sterownik SDN do monitorowania stanu sieci oraz do podejmowania decyzji związanych ze sterowaniem ruchem w sieci. W pracy zbadano możliwości monitorowania sieci w środowisku SDN oraz wiążące się z tym opóźnienia i błędy. Określenie na ile wiarygodne są takie pomiary, może być ważne z punktu widzenia zastosowań techniki SDN w sieciach 5G, np. dla sterowania ruchem w zastosowaniach IoT.
EN
The paper analyses the process of collecting SDN switch statistics. The measurements collected using the OpenFlow protocol are used by the SDN controller to make decisions about the handling of packets and further calculations, e.g. for traffic engineering decisions. This study investigated the possibilities of network monitoring in the SDN environment and the associated packet delays or errors. Determining the reliability of such measurements is important in the use of SDN technique in 5G networks, e.g. in IoT environment.
PL
Pomiary w sieciach SDN mogą być prowadzone i analizowane zarówno przy użyciu metod znanych z tradycyjnych sieci, jak i z wykorzystaniem protokołu interfejsu południowego. Celem niniejszej pracy jest zestawienie rozwiązań prowadzenia pomiarów pasywnych w czasie rzeczywistym. Szczegółowo porównano metody pomiarów obciążenia sieci z wykorzystaniem protokołów OpenFlow i sFlow.
EN
Measurements in SDN can be conducted based on both SDN southband-API and traditional approaches. The aim of this work is to compare real time passive measurement solutions. Comparative analysis of OpenFlow and sFlow protocols was conducted in terms of network load measurements.
EN
This paper presents a new OpenFlow controller: the Distributed Active Information Model (DAIM). The DAIM controller was developed to explore the viability of a logically distributed control plane. It is implemented in a distributed way throughout a software-defined network, at the level of the switches. The method enables local process flows, by way of local packet switching, to be controlled by the distributed DAIM controller (as opposed to a centralised OpenFlow controller). The DAIM ecosystem is discussed with some sample code, together with flowcharts of the implemented algorithms. We present implementation details, a testing methodology, and an experimental evaluation. A performance analysis was conducted using the Cbench open benchmarking tool. Comparisons were drawn with respect to throughput and latency. It is concluded that the DAIM controller can handle a high throughput, while keeping the latency relatively low. We believe the results to date are potentially very interesting, especially in light of the fact that a key feature of the DAIM controller is that it is designed to enable the future development of autonomous local flow process and management strategies.
EN
Distributed Denial of Service (DDoS) flooding attack threats are becoming more and more relevant due to the advances in the Software Defined Networks (SDN). This rising trend creates an emerging need for defense mechanisms against such attacks. In order to address those issues the following paper focuses primarily on the implementation of an automatic real-time DDoS defense application based on sFlow technology. Initially we start by constructing a special flow, bound to a metric in order to capture traffic of interest. As soon as the flow reaches a certain predefined metric level, it is sent to an analyzer. Next we implement a detection algorithm based on the event handling capabilities of the sFlow-RT real-time analyser. Finally, the algorithm is tested with emulation network Mininet using network traffic, resulting in quick and effective DDoS attack mitigation.
PL
Zagrożenia w wyniku rozproszonej odmowy usługi (DDos) w przypadku masowego ataku stają się coraz bardziej możliwe z uwagi na rozwój programowalnych sieci (SDN). Ten rosnący trend powoduje konieczność tworzenia mechanizmów obronnych na wypadek takich ataków. W celu odniesienia sie do takich zagadnień, artykuł ten przede wszystkim skupia się na wdrażaniu automatycznych, działających w czasie rzeczywistym aplikacji obronnych DDoS, opartych na technologii sFlow. Wstępnie rozpoczynamy od stworzenia określonego przepływu, związanego z pomiarem w celu wychwycenia jak duże jest zainteresowanie eksploracją danych. Jak tylko przepływ osiągnie pewien wcześniej określony poziom, informacja zostaje wysłana do analityka. Następnie wdrażamy algorytm wykrywania, w oparciu o zdarzenie, który posiada funkcje analityka sFlow dzialajacego w czasie rzeczywistym. Na koniec testuje się algorytm przy wykorzystaniu emulacyjnej sieci Mininet, wykorzystującej eksplorację danych, co w rezultacie szybko i w efektywny sposób niweluje masowy atak DDoS.
EN
SDN is the approach in telecommunication networks that separates control plane from data forwarding plane by specifying a single network entity as a controller that defines rules (called flows) of traffic forwarding for the switches connected to it. The time that is required for installation of these rules might be a hindrance for the overall performance of SDN network. In the paper, a model for testing and evaluating the influence of traffic payload on flow installation time is presented as well as an implementation of the early experiments’ results are provided.
PL
SDN to rozwiązanie dla sieci telekomunikacyjnych, które separuje płaszczyznę sterowania ruchem od płaszczyzny przesyłania informacji poprzez zastosowanie pojedynczego sterownika jako elementu sieci, który definiuje zasady (nazywane przepływami) kierowania ruchem dla podłączonych do niego switchy. Czas instalacji tych zasad może mieć wpływ na jakość i wydajność całej sieci. W artykule przedstawiono model do testowania i ocenienia wpływu obciążenia ruchem na czas instalacji przepływu oraz opis implementacji i wyniki wstępnych doświadczeń.
EN
Real-time traffic characteristic is different and it is very sensitive to delay. To meet traffic specications in real time, monitoring systems are used as an important part of networking. Many monitoring systems are deployed to have an update view of the network QoS parameters and performance. Most of these systems are implemented to measure QoS parameters in links. Here, in this paper, a system for monitoring queues in each link by means of Software Defined Networks is proposed. The monitoring system is implemented by extending Floodlight controller, which uses OpenFlow as southbound protocol. The controller has a centralized view of the network. By the help of OpenFlow it also can provide flow level statistics. Using these advantages, the proposed system can monitor delay and available bandwidth of a queue on a link or path. Despite of monitoring systems in traditional networks, the proposed monitoring system makes a low overhead in network thanks to OpenFlow protocol messages. It is also integrated into the network controller, which enables QoS and trac engineering applications to use the system's reports for automatic traffic management and QoS setup. The experimental results show a 99% accuracy of the proposed system for monitoring of both bandwidth and delay.
PL
Przedstawiono rozproszone laboratorium SDN, które jest częścią infrastruktury ogólnopolskiej sieci badawczej PL-LAB2020. Laboratorium oferuje heterogeniczne środowisko sprzętowe, które zapewnia szerokie spektrum funkcjonalności dla potencjalnych użytkowników PL-LAB2020.
EN
This article presents the SDN Laboratory which is a part of Polish research infrastructure PL-LAB2020. The laboratory provides distributed and heterogeneous network environment that offers variety of equipment and technologies for Software Defined Networking oriented experiments.
EN
The Software Defined Network (SDN) architecture provides new opportunities to implement security mechanisms in terms of unauthorized activities detection. At the same time, there are certain risks associated with this technology. The presented approach covers a conception of the measurement method, virtual testbed and classification mechanism for SDNs. The paper presents a measurement method which allows collecting network traffic flow parameters, generated by a virtual SDN environment. The collected dataset can be used in machine learning methods to detect unauthorized activities.
EN
The subject of this article is network management through web service calls, which allows software applications to exert an influence on network traffic. In this manner, software can make independent decisions concerning the direction of requests so that they can be served as soon as possible. This is important because only proper cooperation including all architecture layers can ensure the best performance, especially when software that largely depends on computer networks and utilizes them heavily is involved. To demonstrate that the approach described above is feasible and can be useful at the same time, this article presents a switch-level load balancer developed using OpenFlow. Client software communicates with the balancer through REST web service calls, which are used to provide information on current machine load and its ability to serve incoming requests. The result is a cheap, highly customizable and extremely fast load balancer with considerable potential for further development.
PL
Przedstawiono koncepcję sieci sterowanych programowo SDN, zasadność ich stosowania oraz przyszłość. Opisano sieciowe systemy operacyjne wykorzystywane w sieciach SDN oraz środowiska w nich używane. Nacisk położono na środowisko OpenFlow - Podstawowe narzędzie w sieciach sterowanych programowo, a obecnie rozwijane w Katedrze Sieci Telekomunikacyjnych i Komputerowych Politechniki Poznańskiej.
EN
In this article, Software Defined Networks (SDN) as well as need to use them are described. Moreover, future of SDN is mentioned. In addition, network operating systems and environments are introduced. OpenFlow is described in details as the most important solution proposed for SDN. OpenFlow is now improved now by the research group from Chair of Communication and Computer Networks team from Poznan University of Technology within national and European projects.
14
Content available remote Methodology of the direct measurement of the switching latency
EN
The article provides a measurement methodology based on the related RFCs. It brings a solution of determining the switching latency on the physical layer using common measuring devices. As a proof of concept were done a number of the experimental measurements, including analysis of the results. Switching latency is an important performance parameter which participates in the decision-making whether to deploy the switch to low-latency environments. This is especially important in industrial networks for real-time systems involving smart grids. Determine the value of the switching latency is also an important step in the eventual deployment of the OpenFlow technology on this field.
PL
Zaprezentowano metodę badania opóźnień w przełączaniu (switching latency) bazująca na RFC. Parametr ten jest bardzo ważny w sieciach przemysłowych łącznie z sieciami typu smart grid.
EN
This paper describes the performance of various methods of QoS assurance for each connection in an environment composed of virtual networks and dedicate end-to-end connections inside them. The authors worked on the basis of research conducted with the use of the authorial network management system named Executed Management, which uses resources virtualization platforms VMware and Mininet for testing purposes. We briefly describe our system and techniques we used and some alternatives we tested and discarded because of their limitations. Functionality and performance of proposed solution to widespread implemented mechanisms as OpenFlow and MPLS are compared. Reasons for selecting well-known techniques to isolate networks and limit bandwidth on different levels of virtualization are considered. The purpose of this paper is to show out our studies and performance we achieved.
first rewind previous Strona / 1 next fast forward last
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.