Optimizing OpenFlow Load-Balancing with L2 Direct Server Return
|
|
- Dwayne Cain
- 7 years ago
- Views:
Transcription
1 Optimizing OpenFlow Load-Balancing with L2 Direct Server Return Marc Koerner Technische Universitaet Berlin Department of Telecommunication Systems Complex and Distributed IT Systems Einsteinufer 17, Berlin, Germany Odej Kao Technische Universitaet Berlin Department of Telecommunication Systems Complex and Distributed IT Systems Einsteinufer 17, Berlin, Germany Abstract Load-balancers have an essential role in enterprise data-centers, as their existence and performance is decisive for service provision and availability. The state of the art is that today s load-balancers are a piece of specialized embedded hardware with a proprietary operating system and application modules. In this paper we propose a concept to break up this black box model and to show how to develop a high performance LB with OpenFlow technology. In this approach we use OpenFlow network components to process the loadbalancing and in order to reduce the maintenance effort and to lower the hardware cost factor. Furthermore, in case of failover it is redundant by concept and easier to replace by another device. The experimental evaluation of this concept proves that the concept is developed beyond a theoretical concept and the implemented solution is ready for deployment. I. INTRODUCTION Load-balancers have a centralized management and control function in data-centers and ensure the availability of the services that they provide. They distribute incoming service requests to an array of redundant servers to split their workload. This guarantees a fixed minimum amount of available bandwidth and latency for serving all clients. The efficiency increase comes with the problem that load-balancers often represent a single point of failure in a common data-center networking concept. An other problem is that proprietary LB hardware is coupled to a specific administration knowledge regarding the configuration, which is highly vendor depended. To avoid these impacts on configuration and maintenance by keeping the advantages of the devices, the proposed concept avoids the usage of proprietary load-balancing hardware as well as of the proprietary administration interfaces, by using the open standard for software defined networking called OpenFlow. The basic idea of this approach is that OpenFlow standard delivers an interface for performing several forwarding, dropping and modification actions for packets on vendor comprehensive switching and routing hardware. The aim is to show the usage of this new technology and provide business application on this platform. Feasibility and performance values were collected in a local OpenFlow testbed at the Technical University of Berlin. This testbed is part of a European OpenFlow infrastrucutre which is distributed over several European countries. It was created in scope of the project OpenFlow in Europe: Linking Infrastructures and Applications [1]. The remaining of the paper is organized as follows. The following section II introduces all information from the idea and previous investigations up to the required tools for the deployment. Section III explains the architecture and the processing model. Section IV gives an overview about the implementation followed by the evaluation in section V. Finally, section VI summarizes the results and experiences. II. BACKGROUND AND RELATED WORK The OpenFlow [2] paradigm is an approach to create a vendor-independent and a standardized interface for software defined networking. Basically, the idea of OpenFlow standard is to decouple the data- and the control-path of the packet forwarding hardware into two different modules. The switch, which acts as a plain data-path element and the extracted controller, which is completely separated and externally hosted. The protocol provides a manipulation mechanism for the forwarding tables in switches or routers. These forwarding tables are physically realized in ternary content addressable memory (TCAM). OpenFlow provides an abstraction layer for TCAM called flow tables. They contain the flow mods with a packet pattern, an associated actions and a counter. These flow mods are specified by the OpenFlow controller into the flow tables in the data-path element. The concept works with a network of packet processing devices connected to the centralized controller. This means that one single controller is able to control a single switch up to a complete network. OpenFlow controllers can be developed by implementing the specification, which is currently available from version 1.0 [3] up to version 1.3 [4]. Also available are several existing controllers, like NOX [5] or Floodlight [6] with an adaptable application interface (API). This is a comfortable solution for developers because they can focus on the flow processing algorithm or model. NOX is an older controller and has a C++ and Python API. In contrast, Floodlight is a newer controller implementation and is based on the Java Enterprise Edition. FloodLight can be controlled through a Java API. Basically, developers have the opportunity to implement modules or applications working on top of these controllers.
2 Deploying OpenFlow in data-centers can lead to several advantages, from decreasing management complexity up to network and service improvements, e.g. flexibility and scalability [7]. One important data-center element with significant impact on many performance factors is given by load-balancers. Mostly they have a centralized role and also a strong influence on the serving performance of the data-center. Multiple operation modes and balancing algorithms are available. Related papers dealing with OpenFlow based load-balancing focus on HTTP traffic [8], [9] and also does not deliver performance measurements. As observed during former investigations with a destination network address translation (DNAT) method, based on an OpenFlow load-balancing controllers, working with embedded switching hardware [10], there are major performance restrictions. These restrictions belong to the processed operation and the network layer. Especially network address rewrite operations are currently redirected from the application specific integrated circuit (ASIC) to the embedded core of the switch and processed in software. This is the reason for most of the performance issues. This behaviour is currently shown by all hardware OpenFlow switches. Switches are typical layer two devices and most of them used ASIC s or OpenFlow vendor implementations on the switch itself are not able to process layer three manipulation operations with line-rate. A procedure of layer two load-balancing is called direct server return (L2DSR) [11]. The idea of this concept is that a load-balancer in the local server network forwards and distributes service requests and packets directly to an array of servers with the same virtual IP address, equal to the LB network address. The LB solely substitute the destination MAC address and forwards the packet. This mechanism is called MAC address translation (MAT) and works similar as NAT. The server answering the request returns the packets directly to the router in the broadcast domain where they seem to come from, indicated through the layer two source address which is not manipulated by the LB. III. ARCHITECTURE This section introduces the fundamentals how to use existing switch based OpenFlow hardware to deploy a full functional load-balancer which is able to deliver line-rate performance. The main idea of this networking concept is to use the active network components for the load-balancing and thus replace the static LB-Hardware with an OpenFlow controller which balances the load directly on the switching hardware. The focus of this particular approach is a performance concept for load-balancing considering opportunities of OpenFlow. The balancing algorithm is optional replaceable and adoptable to the specific requirements of the service and will not further discussed in this paper. As mentioned in section II a network-layer-three approach has not enough performance for a suitable load-balancing solution as needed by data-centers. So this model uses a layer two concept to improve the processing performance on OpenFlow switching devices. In this architecture, a similar procedure to L2DSR is used for an OpenFlow controller implementation for efficient load balancing. Fig. 1. hardware Load-Balancing concept with L2DSR on OpenFlow switching As depicted in figure 1 the servers are directly connected to an OpenFlow-switch in a routed data-center broadcast domain. The OpenFlow-controller labelled with LBC is responsible for the load-balancing which decouples the servers against the rest of the network in order to avoid address conflicts. As described in section II, L2DSR works with virtual server IP s. This is not needed in this particular case, because the servers are isolated through the OpenFlow-switch or more precisely through the traffic handling of the controller. The controller is not directly forwarding any L2 traffic, he blocks ARP broadcasts and response directly. This means the controller forces the switch to act as a device with an own non-transparent interface in contrast to normal switches. This is realized by handling every kind of address resolution traffic to the legacy network and advertising the associated MAC and IP. This IP is also assigned to every server in the server network array. Internal address resolution requests in the server array network part are directly answered by the switch with the corresponding information of the controller MAT table. All MAC broadcast are not flooded neither in the server network nor in the datacenter network to prevent layer-three address conflicts. If there is no information available, the switch forwards the packet by replacing the MAC address with its own address and answers the request, if it gets the reply form the requested machine. The switch forwards traffic with his associated IP and rewrites the destination MAC address with an address out of the server MAC pool. Which address is used for this process depends on the balancing algorithm. The server reply is also forwarded with a modified source MAC. The switch solely removes the server layer-two source address and replaces it with his own associated one. This ensures a deterministic addressing in the legacy network and an error free operation. The basic procedure is similar to a routing operation. The OpenFlow load-balancer gets packets and forwards them with a new destination data link layer address. The particular difference in that case is that the LB not only separates the broadcast domains, but it also forwards the traffic which is appointed to the server array also using a dedicated loadbalancing algorithm for the traffic distribution among them. The controller is aware of the mapping between the source IP and the destination MAC and uses this knowledge for a statefull packet transfer between client and server due to a fixed flow entry.
3 Src. MAC Dst. MAC Src. IP Dst. IP 1 C n R eth0 C n OF S eth0 2 R eth1 OF S eth0 C n OF S eth0 3 R eth1 S n C n OF S eth0 4 S n R eth1 OF S eth0 C n 5 OF S eth0 R eth1 OF S eth0 C n 6 R eth0 C n OF S eth0 C n TABLE I L2 AND L3 PACKET HEADER MODIFICATION In following we explain the details of packet-processing using the example network depicted in figure 1. Table I shows all header manipulations in the packet on the way through the network, from the arrival at the router, over the load-balancer up to the server and vice versa. An arriving packet (row 1) with the IP of the load-balancer in this particular case the IP, which is associated with the OpenFlow switch by his LBcontroller is forwarded into the local network. Therefore the router replaces the source and the destination address in the layer-two header of the packet (row 2). This corresponds to the standard packet routing operation. The network destination of the packet is the OpenFlow switch (OFS) with his emulated IP and MAC address. The switch forwards the packet to a server in the server network array and replaces the destination MAC address again (row 3). The choice of the MAC address and of the corresponding server depends on the MAC pool as well as on the selected load-balancing algorithm. The destination IP is not replaced, as both addresses are identical. This means that S n and OF S eth0 have the same IP. Also the source IP is not modified because the server directly replies to the requesting device (row 4). The OFS just forwards the packet and replaces the source MAC address. It is again changed to the OFS MAC (row 5). Finally, the router is forwarding the packet by a standard routing operation to the client in the Internet (row 6). IV. IMPLEMENTATION The load-balancer implementation is based on the NOX controller and deployed as a NOX plug-in, which can be loaded during NOX start by using the name of the plug-in as a parameter in the command line interface (CLI). It has no dependencies to other plug-ins and is a stand-alone implementation. The NOX controller uses an event driven model, which calls the corresponding functions of the implemented component container. The plug-in is written in C/C++ and uses the architecture proposed in section III. For sake of simplicity, the initial L2 load-balancing implementation uses a round robin algorithm for distributing the incoming requests to the servers. The packet processing can be separated into two different processing methods. For example, in case of ARP methods the controller directly forces the switch to send out a packet as reaction on a received one. This means, the switch interface acts as an independent device. It is a slow operation, because every interaction has to be approved by the controller. On the other hand, one can use direct forwarding actions with manipulation of the L2 address by OpenFlow actions executed on the switch by an installed FlowMod- Entry. This is a fast operation, because the controller approved the flow only once. Every further packet which matches the FlowMod-Entry is directly modified and forwarded by the switch without additional approval by the controller. The direct forwarding entries consist of a matching pattern and two OpenFlow actions, which are installed in form of a FlowMod in the switch. The first action is the manipulation assignment for the destination MAC address. The second action is the output action which causes the forwarding of the packet. Basically, the important functions of the implementation are responsible for answering ARP requests, pushing down L2 rewrites FlowMod s and handling the load-balancing. Additional functions are used for the management of the LB entries and ARP mappings. The method responsible for the LB matches every incoming new packet except link layer discovery protocol (LLDP) or ARP. If a new incoming flow is detected, the controller directs the flow rules with the rewrite information down to the switch. This procedure requires two FlowMod s per client server mapping which delivers the same scalability regarding FlowMod s as a switching plug-in. The algorithm maps depending on the source IP of the requesting client a server MAC address and stores this information in the internal LB mapping table. The entry is removed, if the corresponding flow expire event from the switch is received. V. EVALUATION For the experimental evaluation we used the local OFELIA OpenFlow testbed at TU Berlin. The testbed consists of three meshed gigabit OpenFlow switches, supporting OpenFlow version 1.0, and three servers. One server for the OFELIA control framework (CF) and two servers for hosting the user VM s. The used virtualization technology is the XEN hypervisor, which provides para-virtualized Debian VM s. These servers for the user VM s provide three experimental gigabit interfaces each, as well as a dedicated control interface. The CF mainly presents to the user a web interface for resource allocation. Resources are VM s as well as a network flowspace. Fig. 2. Physical OFELIA testbed installation at TUB
4 The hardware details of the testbed depicted by figure 2 are the following. The CF is hosted on IBM server with two Intel Xeon quad-core CPU s with 2.4 GHz using a Suse Linux Enterprise Server operating system. The User-VM servers are equipped with an Intel E quad-core CPU with 16GB RAM running a Debian Squeeze. The deployed OpenFlow switches are three NEC IP8800/ S TW and a legacy HP5400 segregation switch as well as an Ixia T1600 testing system with two four port 1GBit line cards and one eight port 100MBit card. Figure 2 also shows the connection between the devices. The black coloured connections are for the management network, the green connections are for the user control network and blue for the OpenFlow experimental network. Only the blue and green networks are accessible and transparent for experimenters or researchers working on the testbed. For testing the implementation on the hardware testbed is only needed a free of charge OFELIA account, which is available on the website. With this account OpenVPN [12] access to OFELIA facility can be used to allocate a VLAN based OpenFlow network slice with the web-based control framework. This slice was connected to the controller, which was also hosted on a VM in the test-bed. For getting comparable results the same experimental installation, as during former research [10], is used in the same configuration. Only the OpenFlow controller implementation was changed and the IP addresses of the server VM s were adapted. flow, the improvement of the old implementation is visible in a considerable performance gain. This increase is caused by using the implemented layer-two concept with packet header rewrite operations which are completely processable on the ASIC. This is also indicated through the CPU consumption of the embedded core. The usage of the switching device core was nearly idle in opposite to the previous layer-three rewrite approach and operation, which were completely processed in software and caused a high core load of approximated 95 percent. The values printed in figure 3 and 4 were collected with iperf and ping application on the Debian 6.0 VM s. They are collected between the client and server VM s to show the performance of the OpenFlow instance in the switch, which connects them. In both diagrams, the blue series is the old DNAT implementation and the red series presents the new layer-two implementation. Fig. 4. Bandwidth Fig. 3. Latency In this experimental evaluation former L3 DNAT measurement results and measurements from the new developed L2 DSR controller and their performance in the data-path are compared with each other. The focus of these measurements is set on the evaluation of processing performance in the datapath and the feasibility for a data-center usage. Except from the flow-mod installation time, where the switch communicates with the controller and asks how to proceed with the As depicted by figure 3 and also summarized in table II, the latency was reduced to a quarter. Also the scatter of the round trip time (RTT) was reduced to a maximum of 0.2ms with a standard deviation of 27ns. While bandwidth with TCP traffic was increased up to 939 MBit per second. This means an increased bandwidth performance by factor 145 and a latency decrease to an average of about 0.5ms. The only exception is the first packet where a new flow have to be approved by the controller. The RTT value for this first packet including the approval and installation of the flow-mod through the controller was measured with 7.62ms. L3DNAT L2DSR Latency: ms ms Bandwidth: 6.46 MBit/s 939 MBit/s TABLE II MEASUREMENT AVERAGE VALUES
5 VI. CONCLUSION This paper presents a novel approach for implementing load-balancing without deployment of dedicated hardware components and solely on the top of an OpenFlow enabled network. We achieve a significant performance gain compared with our previous approaches. The described load-balancing procedure improves the first concept and delivers nearly line-rate bandwidth throughput. With this OpenFlow loadbalancing controller model it is possible to process L2 LB on switching hardware with no limitations regarding performance issues. This result is a further step in the direction of OpenFlow applications for data-centers. When the time has come where software defined networks will replace the normal infrastructure of commercial data-center due to their whole advantages like central management, flexibility, modularity, reliability and so on, this load-balancing controller application is an approach for a real business scenario. The experimental evaluation and the measured real-world results prove that software-defined networking like OpenFlow can cover demands of productive data-centers. The architecture improves an evaluated implementation of a known concept realized with a new technology. REFERENCES [1] Openflow in europe: Linking infrastructure and applications, October 2011, [2] N. McKeown, T. Anderson, H. Balakrishnan, G. Parulkar, L. Peterson, J. Rexford, S. Shenker, and J. Turner, Openflow: Enabling innovation in campus networks, ACM SIGCOMM Computer Communication Review, April [3] T. O. Consortium, Openflow switch specification / version 1.0.0, Decmber 2009, [4] O. N. Foundation, Openflow switch specification / version 1.3.0, May 2012, [5] N. Gude, T. Koponen, J. Pettit, B. Pfaff, M. Casado, N. McKeown, and S. Shenker, Nox: Towards an operating system for networks, ACM SIGCOMM Computer Communication Review, July [6] B. S. Networks, Floodlight, August 2012, openflowhub.org/. [7] A. Tavakoli, M. Casado, T. Koponen, and S. Shenker, Applying nox to the datacenter, Eighth ACM Workshop on Hot Topics in Networks, [8] N. Handigol, S. Seetharaman, M. Flajslik, N. McKeown, and R. Johari, Plug-n-serve: Load-balancing web traffic using openflow, ACM SIGCOMM Computer Communication Review, [9] R. Wang, D. Butnariu, and J. Rexford, Openflow-based server load balancing gonewild, In Hot-ICE, [10] M. Koerner and O. Kao, Multiple service load-balancing with openflow, in Proceedings of the IEEE 13th Conference on High Performance Switching and Routing, ser. IEEE. IEEE publishers, [11] T. Bourke, Server load balancing, O Reilly Media, August [12] Openvpn, September 2012,
Multiple Service Load-Balancing with OpenFlow
2012 IEEE 13th International Conference on High Performance Switching and Routing Multiple Service Load-Balancing with OpenFlow Marc Koerner Technische Universitaet Berlin Department of Telecommunication
More informationComparisons of SDN OpenFlow Controllers over EstiNet: Ryu vs. NOX
Comparisons of SDN OpenFlow Controllers over EstiNet: Ryu vs. NOX Shie-Yuan Wang Hung-Wei Chiu and Chih-Liang Chou Department of Computer Science, National Chiao Tung University, Taiwan Email: shieyuan@cs.nctu.edu.tw
More informationOpenFlow: Enabling Innovation in Campus Networks
OpenFlow: Enabling Innovation in Campus Networks Nick McKeown Stanford University Presenter: Munhwan Choi Table of contents What is OpenFlow? The OpenFlow switch Using OpenFlow OpenFlow Switch Specification
More informationOpenFlow Based Load Balancing
OpenFlow Based Load Balancing Hardeep Uppal and Dane Brandon University of Washington CSE561: Networking Project Report Abstract: In today s high-traffic internet, it is often desirable to have multiple
More informationOpenFlow: Load Balancing in enterprise networks using Floodlight Controller
OpenFlow: Load Balancing in enterprise networks using Floodlight Controller Srinivas Govindraj, Arunkumar Jayaraman, Nitin Khanna, Kaushik Ravi Prakash srinivas.govindraj@colorado.edu, arunkumar.jayaraman@colorado.edu,
More informationTutorial: OpenFlow in GENI
Tutorial: OpenFlow in GENI GENI Project Office The current Internet is at an impasse because new architecture cannot be deployed or even adequately evaluated [PST04] [PST04]: Overcoming the Internet Impasse
More informationImplementation of Address Learning/Packet Forwarding, Firewall and Load Balancing in Floodlight Controller for SDN Network Management
Research Paper Implementation of Address Learning/Packet Forwarding, Firewall and Load Balancing in Floodlight Controller for SDN Network Management Raphael Eweka MSc Student University of East London
More informationSoftware Defined Networking What is it, how does it work, and what is it good for?
Software Defined Networking What is it, how does it work, and what is it good for? slides stolen from Jennifer Rexford, Nick McKeown, Michael Schapira, Scott Shenker, Teemu Koponen, Yotam Harchol and David
More informationXperience of Programmable Network with OpenFlow
International Journal of Computer Theory and Engineering, Vol. 5, No. 2, April 2013 Xperience of Programmable Network with OpenFlow Hasnat Ahmed, Irshad, Muhammad Asif Razzaq, and Adeel Baig each one is
More information基 於 SDN 與 可 程 式 化 硬 體 架 構 之 雲 端 網 路 系 統 交 換 器
基 於 SDN 與 可 程 式 化 硬 體 架 構 之 雲 端 網 路 系 統 交 換 器 楊 竹 星 教 授 國 立 成 功 大 學 電 機 工 程 學 系 Outline Introduction OpenFlow NetFPGA OpenFlow Switch on NetFPGA Development Cases Conclusion 2 Introduction With the proposal
More informationORAN: OpenFlow Routers for Academic Networks
ORAN: OpenFlow Routers for Academic Networks A. Rostami,T.Jungel,A.Koepsel,H.Woesner,A.Wolisz Telecommunication Networks Group (TKN), Technical University of Berlin, Germany {rostami, wolisz}@tkn.tu-berlin.de
More informationOpenFlow and Onix. OpenFlow: Enabling Innovation in Campus Networks. The Problem. We also want. How to run experiments in campus networks?
OpenFlow and Onix Bowei Xu boweixu@umich.edu [1] McKeown et al., "OpenFlow: Enabling Innovation in Campus Networks," ACM SIGCOMM CCR, 38(2):69-74, Apr. 2008. [2] Koponen et al., "Onix: a Distributed Control
More informationA collaborative model for routing in multi-domains OpenFlow networks
A collaborative model for routing in multi-domains OpenFlow networks Xuan Thien Phan, Nam Thoai Faculty of Computer Science and Engineering Ho Chi Minh City University of Technology Ho Chi Minh city, Vietnam
More informationOpenFlow based Load Balancing for Fat-Tree Networks with Multipath Support
OpenFlow based Load Balancing for Fat-Tree Networks with Multipath Support Yu Li and Deng Pan Florida International University Miami, FL Abstract Data center networks are designed for satisfying the data
More informationNetwork performance in virtual infrastructures
Network performance in virtual infrastructures A closer look at Amazon EC2 Alexandru-Dorin GIURGIU University of Amsterdam System and Network Engineering Master 03 February 2010 Coordinators: Paola Grosso
More informationA Study on Software Defined Networking
A Study on Software Defined Networking Yogita Shivaji Hande, M. Akkalakshmi Research Scholar, Dept. of Information Technology, Gitam University, Hyderabad, India Professor, Dept. of Information Technology,
More informationCurrent Trends of Topology Discovery in OpenFlow-based Software Defined Networks
1 Current Trends of Topology Discovery in OpenFlow-based Software Defined Networks Leonardo Ochoa-Aday, Cristina Cervello -Pastor, Member, IEEE, and Adriana Ferna ndez-ferna ndez Abstract The explosion
More informationhp ProLiant network adapter teaming
hp networking june 2003 hp ProLiant network adapter teaming technical white paper table of contents introduction 2 executive summary 2 overview of network addressing 2 layer 2 vs. layer 3 addressing 2
More informationEnabling Fast Failure Recovery in OpenFlow Networks
Enabling Fast Failure Recovery in OpenFlow Networks Sachin Sharma, Dimitri Staessens, Didier Colle, Mario Pickavet and Piet Demeester Ghent University - IBBT, Department of Information Technology (INTEC),
More informationInformation- Centric Networks. Section # 13.2: Alternatives Instructor: George Xylomenos Department: Informatics
Information- Centric Networks Section # 13.2: Alternatives Instructor: George Xylomenos Department: Informatics Funding These educational materials have been developed as part of the instructors educational
More informationSteroid OpenFlow Service: Seamless Network Service Delivery in Software Defined Networks
Steroid OpenFlow Service: Seamless Network Service Delivery in Software Defined Networks Aaron Rosen and Kuang-Ching Wang Holcombe Department of Electrical and Computer Engineering Clemson University Clemson,
More informationEthernet-based Software Defined Network (SDN) Cloud Computing Research Center for Mobile Applications (CCMA), ITRI 雲 端 運 算 行 動 應 用 研 究 中 心
Ethernet-based Software Defined Network (SDN) Cloud Computing Research Center for Mobile Applications (CCMA), ITRI 雲 端 運 算 行 動 應 用 研 究 中 心 1 SDN Introduction Decoupling of control plane from data plane
More informationZEN LOAD BALANCER EE v3.04 DATASHEET The Load Balancing made easy
ZEN LOAD BALANCER EE v3.04 DATASHEET The Load Balancing made easy OVERVIEW The global communication and the continuous growth of services provided through the Internet or local infrastructure require to
More informationOpen Source Network: Software-Defined Networking (SDN) and OpenFlow
Open Source Network: Software-Defined Networking (SDN) and OpenFlow Insop Song, Ericsson LinuxCon North America, Aug. 2012, San Diego CA Objectives Overview of OpenFlow Overview of Software Defined Networking
More informationLimitations of Current Networking Architecture OpenFlow Architecture
CECS 572 Student Name Monday/Wednesday 5:00 PM Dr. Tracy Bradley Maples OpenFlow OpenFlow is the first open standard communications interface that enables Software Defined Networking (SDN) [6]. It was
More informationSoftware Defined Networks (SDN)
Software Defined Networks (SDN) Nick McKeown Stanford University With: Martín Casado, Teemu Koponen, Scott Shenker and many others With thanks to: NSF, GPO, Stanford Clean Slate Program, Cisco, DoCoMo,
More informationSDN. What's Software Defined Networking? Angelo Capossele
SDN What's Software Defined Networking? Angelo Capossele Outline Introduction to SDN OpenFlow Network Functions Virtualization Some examples Opportunities Research problems Security Case study: LTE (Mini)Tutorial
More informationOutline. Institute of Computer and Communication Network Engineering. Institute of Computer and Communication Network Engineering
Institute of Computer and Communication Network Engineering Institute of Computer and Communication Network Engineering Communication Networks Software Defined Networking (SDN) Prof. Dr. Admela Jukan Dr.
More informationData Analysis Load Balancer
Data Analysis Load Balancer Design Document: Version: 1.0 Last saved by Chris Small April 12, 2010 Abstract: The project is to design a mechanism to load balance network traffic over multiple different
More informationSoftware Defined Networks
Software Defined Networks Damiano Carra Università degli Studi di Verona Dipartimento di Informatica Acknowledgements! Credits Part of the course material is based on slides provided by the following authors
More informationSoftware Defined Networking What is it, how does it work, and what is it good for?
Software Defined Networking What is it, how does it work, and what is it good for? Many slides stolen from Jennifer Rexford, Nick McKeown, Scott Shenker, Teemu Koponen, Yotam Harchol and David Hay Agenda
More information50. DFN Betriebstagung
50. DFN Betriebstagung IPS Serial Clustering in 10GbE Environment Tuukka Helander, Stonesoft Germany GmbH Frank Brüggemann, RWTH Aachen Slide 1 Agenda Introduction Stonesoft clustering Firewall parallel
More informationZEN LOAD BALANCER EE v3.02 DATASHEET The Load Balancing made easy
ZEN LOAD BALANCER EE v3.02 DATASHEET The Load Balancing made easy OVERVIEW The global communication and the continuous growth of services provided through the Internet or local infrastructure require to
More informationHow To Understand The Power Of The Internet
DATA COMMUNICATOIN NETWORKING Instructor: Ouldooz Baghban Karimi Course Book: Computer Networking, A Top-Down Approach, Kurose, Ross Slides: - Course book Slides - Slides from Princeton University COS461
More informationProgrammable Networking with Open vswitch
Programmable Networking with Open vswitch Jesse Gross LinuxCon September, 2013 2009 VMware Inc. All rights reserved Background: The Evolution of Data Centers Virtualization has created data center workloads
More informationSecuring Local Area Network with OpenFlow
Securing Local Area Network with OpenFlow Master s Thesis Presentation Fahad B. H. Chowdhury Supervisor: Professor Jukka Manner Advisor: Timo Kiravuo Department of Communications and Networking Aalto University
More informationAUTO DEFAULT GATEWAY SETTINGS FOR VIRTUAL MACHINES IN SERVERS USING DEFAULT GATEWAY WEIGHT SETTINGS PROTOCOL (DGW)
AUTO DEFAULT GATEWAY SETTINGS FOR VIRTUAL MACHINES IN SERVERS USING DEFAULT GATEWAY WEIGHT SETTINGS PROTOCOL (DGW) Suman Dutta 1, Shouman Barua 2 and Jishu Sen 3 1 IT Trainer, Logitrain.com.au 2 PhD research
More informationAutonomicity Design in OpenFlow Based Software Defined Networking
GC'12 Workshop: The 4th IEEE International Workshop on Management of Emerging Networks and Services Autonomicity Design in OpenFlow Based Software Defined Networking WANG Wendong, Yannan HU, Xirong QUE,
More informationHow SDN will shape networking
How SDN will shape networking Nick McKeown Stanford University With: Martín Casado, Teemu Koponen, Sco> Shenker and many others With thanks to: NSF, GPO, Stanford Clean Slate Program, Cisco, DoCoMo, DT,
More informationSOFTWARE-DEFINED NETWORKING AND OPENFLOW
SOFTWARE-DEFINED NETWORKING AND OPENFLOW Freddie Örnebjär TREX Workshop 2012 2012 Brocade Communications Systems, Inc. 2012/09/14 Software-Defined Networking (SDN): Fundamental Control
More informationSoftware Defined Networking Basics
Software Defined Networking Basics Anupama Potluri School of Computer and Information Sciences University of Hyderabad Software Defined Networking (SDN) is considered as a paradigm shift in how networking
More informationWedge Networks: Transparent Service Insertion in SDNs Using OpenFlow
Wedge Networks: EXECUTIVE SUMMARY In this paper, we will describe a novel way to insert Wedge Network s multiple content security services (such as Anti-Virus, Anti-Spam, Web Filtering, Data Loss Prevention,
More informationAxon: A Flexible Substrate for Source- routed Ethernet. Jeffrey Shafer Brent Stephens Michael Foss Sco6 Rixner Alan L. Cox
Axon: A Flexible Substrate for Source- routed Ethernet Jeffrey Shafer Brent Stephens Michael Foss Sco6 Rixner Alan L. Cox 2 Ethernet Tradeoffs Strengths Weaknesses Cheap Simple High data rate Ubiquitous
More informationResearch Article Dynamic Server Cluster Load Balancing in Virtualization Environment with OpenFlow
International Journal of Distributed Sensor Networks Volume 215, Article ID 531538, 9 pages http://dx.doi.org/1.1155/215/531538 Research Article Dynamic Server Cluster Load Balancing in Virtualization
More informationOpenFlow: Concept and Practice. Dukhyun Chang (dhchang@mmlab.snu.ac.kr)
OpenFlow: Concept and Practice Dukhyun Chang (dhchang@mmlab.snu.ac.kr) 1 Contents Software-Defined Networking (SDN) Overview of OpenFlow Experiment with OpenFlow 2/24 Software Defined Networking.. decoupling
More informationSoftware Defined Networking and the design of OpenFlow switches
Software Defined Networking and the design of OpenFlow switches Paolo Giaccone Notes for the class on Packet Switch Architectures Politecnico di Torino December 2015 Outline 1 Introduction to SDN 2 OpenFlow
More informationSDN CENTRALIZED NETWORK COMMAND AND CONTROL
SDN CENTRALIZED NETWORK COMMAND AND CONTROL Software Defined Networking (SDN) is a hot topic in the data center and cloud community. The geniuses over at IDC predict a $2 billion market by 2016
More informationIMPLEMENTATION AND EVALUATION OF THE MOBILITYFIRST PROTOCOL STACK ON SOFTWARE-DEFINED NETWORK PLATFORMS
IMPLEMENTATION AND EVALUATION OF THE MOBILITYFIRST PROTOCOL STACK ON SOFTWARE-DEFINED NETWORK PLATFORMS BY ARAVIND KRISHNAMOORTHY A thesis submitted to the Graduate School New Brunswick Rutgers, The State
More informationOpenFlow Overview. Daniel Turull danieltt@kth.se
OpenFlow Overview Daniel Turull danieltt@kth.se Overview OpenFlow Software Defined Networks (SDN) Network Systems Lab activities Daniel Turull - Netnod spring meeting 2012 2 OpenFlow Why and where was
More informationGetting to know OpenFlow. Nick Rutherford Mariano Vallés {nicholas,mariano}@ac.upc.edu
Getting to know OpenFlow Nick Rutherford Mariano Vallés {nicholas,mariano}@ac.upc.edu OpenFlow Switching 1. A way to run experiments in the networks we use everyday. A pragmatic compromise Allow researchers
More informationA Testbed for research and development of SDN applications using OpenFlow
A Testbed for research and development of SDN applications using OpenFlow Nádia Pires Gonçalves nadia.goncalves@tecnico.ulisboa.pt Instituto Superior Técnico, Universidade de Lisboa ABSTRACT Network technologies
More informationOpenFlow-Based Dynamic Server Cluster Load Balancing with Measurement Support
OpenFlow-Based Dynamic Server Cluster Load Balancing with Measurement Support Qingwei Du and Huaidong Zhuang College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics,
More informationWhat is VLAN Routing?
Application Note #38 February 2004 What is VLAN Routing? This Application Notes relates to the following Dell product(s): 6024 and 6024F 33xx Abstract Virtual LANs (VLANs) offer a method of dividing one
More informationWHITE PAPER. SDN Controller Testing: Part 1
WHITE PAPER SDN Controller Testing: Part 1 www.ixiacom.com 915-0946-01 Rev. A, April 2014 2 Table of Contents Introduction... 4 Testing SDN... 5 Methodologies... 6 Testing OpenFlow Network Topology Discovery...
More informationNetwork Virtualization and Data Center Networks 263-3825-00 Data Center Virtualization - Basics. Qin Yin Fall Semester 2013
Network Virtualization and Data Center Networks 263-3825-00 Data Center Virtualization - Basics Qin Yin Fall Semester 2013 1 Walmart s Data Center 2 Amadeus Data Center 3 Google s Data Center 4 Data Center
More informationIntel Ethernet Switch Load Balancing System Design Using Advanced Features in Intel Ethernet Switch Family
Intel Ethernet Switch Load Balancing System Design Using Advanced Features in Intel Ethernet Switch Family White Paper June, 2008 Legal INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL
More informationSDN, OpenFlow and the ONF
SDN, OpenFlow and the ONF OpenFlow/Software-Defined Networking (SDN) OpenFlow/SDN is emerging as one of the most promising and disruptive networking technologies of recent years. It has the potential to
More informationPerformance of Host Identity Protocol on Nokia Internet Tablet
Performance of Host Identity Protocol on Nokia Internet Tablet Andrey Khurri Helsinki Institute for Information Technology HIP Research Group IETF 68 Prague March 23, 2007
More informationulobal: Enabling In-Network Load Balancing for Arbitrary Internet Services on SDN
ulobal: Enabling In-Network Load Balancing for Arbitrary Internet Services on SDN Alex F R Trajano, Marcial P Fernandez Universidade Estadual do Ceará Fortaleza, Ceará, Brazil Email: alex.ferreira@uece.br,
More informationVirtualization Technologies (ENCS 691K Chapter 3)
Virtualization Technologies (ENCS 691K Chapter 3) Roch Glitho, PhD Associate Professor and Canada Research Chair My URL - http://users.encs.concordia.ca/~glitho/ The Key Technologies on Which Cloud Computing
More informationPOX CONTROLLER PERFORMANCE FOR OPENFLOW NETWORKS. Selçuk Yazar, Erdem Uçar POX CONTROLLER ЗА OPENFLOW ПЛАТФОРМА. Селчук Язар, Ердем Учар
УПРАВЛЕНИЕ И ОБРАЗОВАНИЕ MANAGEMENT AND EDUCATION TOM IX (6) 2013 VOL. IX (6) 2013 POX CONTROLLER PERFORMANCE FOR OPENFLOW NETWORKS Selçuk Yazar, Erdem Uçar POX CONTROLLER ЗА OPENFLOW ПЛАТФОРМА Селчук
More informationDesign and Implementation of Dynamic load balancer on OpenFlow enabled SDNs
IOSR Journal of Engineering (IOSRJEN) e-issn: 2250-3021, p-issn: 2278-8719 Vol. 3, Issue 8 (August. 2013), V4 PP 32-41 Design and Implementation of Dynamic load balancer on OpenFlow enabled SDNs Ragalatha
More informationNetwork Virtualization Based on Flows
TERENA NETWORKING CONFERENCE 2009 June 9, 2009 Network Virtualization Based on Flows Peter Sjödin Markus Hidell, Georgia Kontesidou, Kyriakos Zarifis KTH Royal Institute of Technology, Stockholm Outline
More informationlibnetvirt: the network virtualization library
libnetvirt: the network virtualization library Daniel Turull, Markus Hidell, Peter Sjödin KTH Royal Institute of Technology, School of ICT Stockholm, Sweden Email: {danieltt,mahidell,psj}@kth.se Abstract
More informationSecurity Challenges & Opportunities in Software Defined Networks (SDN)
Security Challenges & Opportunities in Software Defined Networks (SDN) June 30 th, 2015 SEC2 2015 Premier atelier sur la sécurité dans les Clouds Nizar KHEIR Cyber Security Researcher Orange Labs Products
More informationAdvanced Study of SDN/OpenFlow controllers
Advanced Study of SDN/OpenFlow controllers Alexander Shalimov ashalimov@arccn.ru Vasily Pashkov vpashkov@arccn.ru Dmitry Zuikov dzuikov@arccn.ru Ruslan Smeliansky smel@arccn.ru Daria Zimarina zimarina@lvk.cs.msu.su
More informationSDN. WHITE PAPER Intel Ethernet Switch FM6000 Series - Software Defined Networking. Recep Ozdag Intel Corporation
WHITE PAPER Intel Ethernet Switch FM6000 Series - Software Defined Networking Intel Ethernet Switch FM6000 Series - Software Defined Networking Recep Ozdag Intel Corporation Software Defined Networking
More information642 523 Securing Networks with PIX and ASA
642 523 Securing Networks with PIX and ASA Course Number: 642 523 Length: 1 Day(s) Course Overview This course is part of the training for the Cisco Certified Security Professional and the Cisco Firewall
More informationExtensible and Scalable Network Monitoring Using OpenSAFE
Extensible and Scalable Network Monitoring Using OpenSAFE Jeffrey R. Ballard ballard@cs.wisc.edu Ian Rae ian@cs.wisc.edu Aditya Akella akella@cs.wisc.edu Abstract Administrators of today s networks are
More informationDesign and implementation of server cluster dynamic load balancing in virtualization environment based on OpenFlow
Design and implementation of server cluster dynamic load balancing in virtualization environment based on OpenFlow Wenbo Chen Hui Li Qiang Ma Zhihao Shang Lanzhou University Lanzhou University Lanzhou
More informationCCNA R&S: Introduction to Networks. Chapter 5: Ethernet
CCNA R&S: Introduction to Networks Chapter 5: Ethernet 5.0.1.1 Introduction The OSI physical layer provides the means to transport the bits that make up a data link layer frame across the network media.
More informationScalable Network Virtualization in Software-Defined Networks
Scalable Network Virtualization in Software-Defined Networks Dmitry Drutskoy Princeton University Eric Keller University of Pennsylvania Jennifer Rexford Princeton University ABSTRACT Network virtualization
More informationOFCProbe: A Platform-Independent Tool for OpenFlow Controller Analysis
c 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising
More informationSoftware Defined Networking and OpenFlow: a Concise Review
Software Defined Networking and OpenFlow: a Concise Review Stefano Forti stefano.forti92@gmail.com MSc in Computer Science and Networking Scuola Superiore Sant'Anna - University of Pisa 1. Introduction
More informationThe Internet: A Remarkable Story. Inside the Net: A Different Story. Networks are Hard to Manage. Software Defined Networking Concepts
The Internet: A Remarkable Story Software Defined Networking Concepts Based on the materials from Jennifer Rexford (Princeton) and Nick McKeown(Stanford) Tremendous success From research experiment to
More informationOpenFlow. Ihsan Ayyub Qazi. Slides use info from Nick Mckeown
OpenFlow Ihsan Ayyub Qazi Slides use info from Nick Mckeown Why Openflow? Deployment of new experimental protocols is challenging Huge installed base of protocols and equipment Reluctance by network operators
More informationDNS ROUND ROBIN HIGH-AVAILABILITY LOAD SHARING
PolyServe High-Availability Server Clustering for E-Business 918 Parker Street Berkeley, California 94710 (510) 665-2929 wwwpolyservecom Number 990903 WHITE PAPER DNS ROUND ROBIN HIGH-AVAILABILITY LOAD
More informationNetwork Virtualization and Software-defined Networking. Chris Wright and Thomas Graf Red Hat June 14, 2013
Network Virtualization and Software-defined Networking Chris Wright and Thomas Graf Red Hat June 14, 2013 Agenda Problem Statement Definitions Solutions She can't take much more of this, captain! Challenges
More informationProtecting and controlling Virtual LANs by Linux router-firewall
Protecting and controlling Virtual LANs by Linux router-firewall Tihomir Katić Mile Šikić Krešimir Šikić Faculty of Electrical Engineering and Computing University of Zagreb Unska 3, HR 10000 Zagreb, Croatia
More informationSoftware Defined Networking
Software Defined Networking Stefano Giordano Dipartimento di Ingegneria dell Informazione Università di Pisa 3D Reference model of ISDN Hourglass reference model of a TCP/IP network Network Ossification
More informationAuto-Configuration of SDN Switches in SDN/Non-SDN Hybrid Network
Auto-Configuration of SDN Switches in SDN/Non-SDN Hybrid Network Rohit Katiyar cs13m1016@iith.ac.in Prakash Pawar cs13m1015@iith.ac.in Kotaro Kataoka kotaro@iith.ac.in Abhay Gupta cs12b1041@iith.ac.in
More informationNETWORK MANAGEMENT SERVICES BASED ON THE OPENFLOW ENVIRONMENT
Computer Science 15 (3) 2014 http://dx.doi.org/10.7494/csci.2014.15.3.253 Paweł Wilk Piotr Nawrocki NETWORK MANAGEMENT SERVICES BASED ON THE OPENFLOW ENVIRONMENT Abstract The subject of this article is
More informationStatus of OpenFlow research and test facilities in Europe
Status of OpenFlow research and test facilities in Europe Cape Town, Nov 14-15, 2011 Dr.-Ing. Hagen Woesner EICT GmbH, Berlin, Germany Agenda Software Defined Networks and OpenFlow Architectural Splits
More informationEnabling Software Defined Networking using OpenFlow
Enabling Software Defined Networking using OpenFlow 1 Karamjeet Kaur, 2 Sukhveer Kaur, 3 Vipin Gupta 1,2 SBS State Technical Campus Ferozepur, 3 U-Net Solutions Moga Abstract Software Defined Networking
More informationSOFTWARE-DEFINED NETWORKING AND OPENFLOW
SOFTWARE-DEFINED NETWORKING AND OPENFLOW Eric Choi < echoi@brocade.com> Senior Manager, Service Provider Business Unit, APJ 2012 Brocade Communications Systems, Inc. EPF 7 2012/09/17 Software-Defined Networking
More informationHow To Compare Performance Of A Router On A Hypervisor On A Linux Virtualbox 2.5 (Xen) To A Virtualbox 3.5.2 (Xeen) 2.2.5-Xen-Virtualization (X
Performance Evaluation of Virtual Routers in Para-virtual Environment 1. Abhishek Bajaj abhishek.bajaj@iiitb.net 2. Anargha Biswas anargha.biswas@iiitb.net 3. Ambarish Kumar ambarish.kumar@iiitb.net 4.
More informationOVERLAYING VIRTUALIZED LAYER 2 NETWORKS OVER LAYER 3 NETWORKS
OVERLAYING VIRTUALIZED LAYER 2 NETWORKS OVER LAYER 3 NETWORKS Matt Eclavea (meclavea@brocade.com) Senior Solutions Architect, Brocade Communications Inc. Jim Allen (jallen@llnw.com) Senior Architect, Limelight
More informationCloud Networking Disruption with Software Defined Network Virtualization. Ali Khayam
Cloud Networking Disruption with Software Defined Network Virtualization Ali Khayam In the next one hour Let s discuss two disruptive new paradigms in the world of networking: Network Virtualization Software
More informationVXLAN Performance Evaluation on VMware vsphere 5.1
VXLAN Performance Evaluation on VMware vsphere 5.1 Performance Study TECHNICAL WHITEPAPER Table of Contents Introduction... 3 VXLAN Performance Considerations... 3 Test Configuration... 4 Results... 5
More informationInternet Protocol: IP packet headers. vendredi 18 octobre 13
Internet Protocol: IP packet headers 1 IPv4 header V L TOS Total Length Identification F Frag TTL Proto Checksum Options Source address Destination address Data (payload) Padding V: Version (IPv4 ; IPv6)
More informationViSION Status Update. Dan Savu Stefan Stancu. D. Savu - CERN openlab
ViSION Status Update Dan Savu Stefan Stancu D. Savu - CERN openlab 1 Overview Introduction Update on Software Defined Networking ViSION Software Stack HP SDN Controller ViSION Core Framework Load Balancer
More informationQuantifying the Performance Degradation of IPv6 for TCP in Windows and Linux Networking
Quantifying the Performance Degradation of IPv6 for TCP in Windows and Linux Networking Burjiz Soorty School of Computing and Mathematical Sciences Auckland University of Technology Auckland, New Zealand
More informationSoftware Defined Networking Architecture
Software Defined Networking Architecture Brighten Godfrey CS 538 October 8 2013 slides 2010-2013 by Brighten Godfrey The Problem Networks are complicated Just like any computer system Worse: it s distributed
More informationDeployment Guide. How to prepare your environment for an OnApp Cloud deployment.
Deployment Guide How to prepare your environment for an OnApp Cloud deployment. Document version 1.07 Document release date 28 th November 2011 document revisions 1 Contents 1. Overview... 3 2. Network
More informationReal-World Insights from an SDN Lab. Ron Milford Manager, InCNTRE SDN Lab Indiana University
Real-World Insights from an SDN Lab Ron Milford Manager, InCNTRE SDN Lab Indiana University 1 A bit about IU, the GlobalNOC, and InCNTRE... Indiana University s Network History 1998 University Corporation
More informationAvailability Digest. www.availabilitydigest.com. Redundant Load Balancing for High Availability July 2013
the Availability Digest Redundant Load Balancing for High Availability July 2013 A large data center can comprise hundreds or thousands of servers. These servers must not only be interconnected, but they
More informationOpenFlow: Enabling Innovation in Campus Networks
: Enabling Innovation in Campus Networks March 14, 2008 Nick McKeown Stanford University Guru Parulkar Stanford University Scott Shenker University of California, Berkeley Tom Anderson University of Washington
More informationSoftware Defined Networking (SDN) - Open Flow
Software Defined Networking (SDN) - Open Flow Introduction Current Internet: egalitarian routing/delivery based on destination address, best effort. Future Internet: criteria based traffic management,
More informationHow To Understand The Power Of A Network In A Microsoft Computer System (For A Micronetworking)
Digitaalne andmeülekanne IRT0150 OpenFlow /nädal 6/ Avo Ots avo.ots@ttu.ee 12. märts 2015 1 Various Services Virtual Networks LINP1 LINP3 LINP2 LINP1 Manager LINP2 Manager LINP3 Manager Virtual Resources
More informationNetwork Virtualization History. Network Virtualization History. Extending networking into the virtualization layer. Problem: Isolation
Network irtualization History Network irtualization and Data Center Networks 263-3825-00 SDN Network irtualization Qin Yin Fall Semester 203 Reference: The Past, Present, and Future of Software Defined
More information