Matteo Gerola

ONOS: towards an open, distributed SDN OS

We present our experiences to date building ONOS (Open Network Operating System), an experimental distributed SDN control platform motivated by the performance, scalability, and availability requirements of large operator networks. We describe and evaluate two ONOS prototypes. The first version implemented core features: a distributed, but logically centralized, global network view; scale-out; and fault tolerance. The second version focused on improving performance. Based on experience with these prototypes, we identify additional steps that will be required for ONOS to support use cases such as core network traffic engineering and scheduling, and to become a usable open source, distributed network OS platform that the SDN community can build upon.

VeRTIGO: Network Virtualization and Beyond

In this paper we present Vertigo (Virtual Topologies Generalization in OpenFlow networks), a Software-defined networking platform designed for network virtualization. Based on the OpenFlow original network slicing system Flow Visor, the Vertigo platform aims at covering all flavors of network virtualization: in particular, it is able to expose a simple abstract node on one extreme, and to deliver a logically fully connected network at the very opposite end. In this work, we first introduce the Vertigo system architecture and its design choices, then we report on a prototypical implementation deployed over an OpenFlow-enabled test bed. Experimental results show that Vertigo can deliver flexible and reliable network virtualization services to a wide range of use cases in spite of failure and/or congestion at the underlying physical network.

Generalizing Virtual Network Topologies in OpenFlow-Based Networks

Network Virtualization (NV) is one of the most promising technique to enable innovation in todays network. A recent approach toward NV has been proposed through FlowVisor, whose aim is to leverage on the specific features of an OpenFlow-controlled network to share the same hardware forwarding plane among multiple logical networks. However, FlowVisor lacks some features to enable a full implementation of a NV architecture: the virtual topologies that can be established are restricted to subsets of the physical topology and it has no way for two slices to share flowspace and simultaneously prevent them from interfering with each others traffic. In this work, an innovative system called ADVisor (ADvanced FlowVisor) which enhances FlowVisor while overcoming its major constraints is presented and a set of experimental results discussed to demonstrate its capability to provide an effective support toward a Network Virtualization architecture.

Design and implementation of an OpenFlow hardware abstraction layer

OpenFlow is a leading standard for Software-Defined Networking (SDN) and has already played a significant role in reshaping network infrastructures. However, a wide range of existing provider domains is still not equipped with a framework that supports wider deployment of an OpenFlow-based control plane beyond Ethernet-dominated networks. We address this gap by introducing a Hardware Abstraction Layer (HAL) which can transform legacy network elements into OpenFlow capable devices. This paper details the functional architecture of HAL, discusses the key design aspects and explains how HAL can support a number of network device classes. In addition, this paper presents the implementation details of HAL for hardware platforms such as DOCSIS (Data over Cable Service Interface Specification) and DWDM (Dense Wavelength Division Multiplexing) which have so far received little attention by the OpenFlow research community despite their wide real-world deployment.

Demonstrating generalized virtual topologies in an openflow network

Network Virtualization (NV) is one of the most promising approaches to enable innovation in today’s network. Generally speaking, NV refers to the possibility of pooling together low–level hardware and software resources belonging to a networked system into a single administrative entity. In such a way network resources could be effectively shared in a transparent way among different logical network instances corresponding to different virtual network topologies. A recent approach toward Network Virtualization has been proposed through FlowVisor [1], whose aim is to leverage on the specific features of an OpenFlow–controlled network [2] to share the same hardware forwarding plane among multiple logical networks. As highlighted by the authors in [1], one of the major limitations of FlowVisor is the inability to establish virtual topologies not restricted by the underpinning physical topology. As a consequence, FlowVisor is unable to provide researchers flexibility in designing their experiments with arbitrary network topologies on a defined physical infrastructure. The architecture presented in Chapter 2 of this paper, called ADVisor (ADvanced FlowVisor), provides the functionalities to overcome the above-mentioned FlowVisor’s limitation by allowing the instantiation of generalized virtual topologies in a OpenFlow network through the implementation of virtual links as aggregation of multiple physical links and OpenFlow-enabled switches. In this demo we will show the configuration of a simple virtual topology performed through a Web-based control framework which allows the reservation of network resources (nodes, links and bandwidth) and the management of virtual resources (virtual links and virtual ports). We will also demonstrate the effective instantiation of the virtual topology by running a synthetic traffic generator application.

A Datapath-Centric Virtualization Mechanism for OpenFlow Networks

The adoption of a robust and scalable network virtualization framework is a key requirement in order to make the vision of a shareable network infrastructure a reality. To this aim, one of the most suitable approaches is the one which takes advantage of the emerging paradigm of Software-Defined Networking (SDN) and OpenFlow, its de-facto standard. Several virtualization frameworks have been proposed in the last few years, however, they are either based on proxy-based solutions that raises scalability and robustness issues (FlowVisor), or they rely on a simplified view of the data path (generally based on Open vSwitch instances) that have little chances to be adopted in production network settings. This paper presents a novel OpenFlow-based network virtualization mechanism exploiting a recent open-source data path project named extensible Data path Daemon (xDPd), the proposed multi-platform data path is based on a robust distributed virtualization architecture that is able to run on multi-version OpenFlow switch network scenarios, has a minimal overhead from a performance point of view and can be easily ported on several hardware platforms via xDPd libraries.

Hybrid IP/SDN Networking: Open Implementation and Experiment Management Tools

The introduction of SDN in large-scale IP provider networks is still an open issue and different solutions have been suggested so far. In this paper, we propose a hybrid approach that allows the coexistence of traditional IP routing with SDN based forwarding within the same provider domain. The solution is called OSHI-Open Source Hybrid IP/SDN networking, as we have fully implemented it combining and extending open source software. We discuss the OSHI system architecture and the design and implementation of advanced services like pseudo wires and virtual switches. In addition, we describe a set of open source management tools for the emulation of the proposed solution using either the Mininet emulator or distributed physical testbeds. We refer to this suite of tools as Mantoo (management tools). Mantoo includes an extensible Web-based graphical topology designer, which provides different layered network “views” (e.g., from physical links to service relationships among nodes). The suite can validate an input topology, automatically deploy it over a Mininet emulator or a distributed SDN testbed and allows access to emulated nodes by opening consoles in the web GUI. Mantoo provides also tools to evaluate the performance of the deployed nodes.The introduction of SDN in large-scale IP provider networks is still an open issue and different solutions have been suggested so far. In this paper we propose a hybrid approach allowing the coexistence of traditional IP routing with SDN based forwarding within the same provider domain. The solution is called OSHI – Open Source Hybrid IP/SDN networking as we have fully implemented it combining and extending Open Source software. We discuss the OSHI system architecture and the design and implementation of advanced services like Pseudo Wires and Virtual Switches. Moreover we designed and implemented a set of Open Source management tools for the emulation of the proposed solution over the Mininet emulator and over distributed testbeds. We refer to this suite of tools as Mantoo (Management tools). Mantoo includes an extensible webbased graphical topology designer providing different layered network “views” (e.g. from physical links to service relationships among nodes). The framework is able to validate a topology, to automatically deploy it over a Mininet emulator or a distributed SDN testbed, to access nodes by opening consoles directly via the web GUI. Finally, Mantoo provides tools for evaluating the performance of the deployed solution.

ICONA: Inter Cluster Onos Network application

Several Network Operating Systems have been proposed in the last few years for Software Defined Networks; however, only few of them are offering resiliency, scalability and high availability required for production environments. In our demonstration we present a geographically distributed SDN Control Plane, called ICONA, build on top of the Open Networking Operating System (ONOS) and designed to meet the aforementioned Service Providers requirements. During the demo, that runs inside the GEANT OpenFlow pan-european testbed, we show how a Service Provider engineer can easily manage and monitor the network, deploy some services and how ICONA can automatically recover from Control and Data planes failures.

A distributed network monitoring framework for wireless networks

Multi-hop wireless networks are emerging as a viable alternative for building access networks in areas where conventional solutions (cellular, fiber) are neither feasible nor attractive from an economical standpoint. The management of such networks represents an overly complex task because of the time-varying nature of the radio channel, the mobility of users and the presence of adaptive, self-configuration features. Various solutions are currently being researched, whereby network management functionalities are performed autonomously at the network nodes themselves. Such approaches require a monitoring framework able to bring network-level information to the relevant decision points in an effective and robust manner. In this paper, we present a distributed network monitoring framework, specifically developed for wireless multi-hop networks. The system architecture and the protocols are presented together with results obtained using a prototypical implementation over a real-world testbed. Experimental results show that the framework generates a limited amount of traffic, and that the system is able to consistently recover from node failures.

Demonstrating a Distributed and Version-Agnostic OpenFlow Slicing Mechanism

Several virtualization frameworks have been proposed in the last few years for Software Defined Networks (SDN), however, they are either based on proxy-based solutions that raises scalability and robustness issues (FlowVisor), or they leverage on a simplified view of the data-path (generally based on Open vSwitch instances) that have little chances to be adopted in production network settings. In our demonstration we present preliminary results obtained by deploying and using a novel OpenFlow-based network virtualization mechanism. The mechanism is based on a recently proposed distributed virtualization architecture that is able to run on multi-version OpenFlow scenarios.