Section: New Results

Experimental Environment for Future Internet Architecture

Participants : Walid Dabbous, Thierry Parmentelat, Fréderic Urbani, Daniel Camara, Alina Quereilhac, Shafqat Ur-Rehman, Mohamed Larabi, Thierry Turletti, Julien Tribino.

• SFA Federation of experimental testbeds

  We are now involved in the NOVI (E.U. STREP) project, the F-Lab (French A.N.R.) project, the FED4FIRE (E.U. IP) project and have the lead of the “Control Plane Extensions” WorkPackage of OpenLab (E.U. IP) project. Within these frameworks, as part of the co-development agreement between the Planète team and Princeton University, we have made a great deal of contributions into one of the most visible and renown implementations of the Testbed-Federation architecture known as SFA for Slice-based Federation Architecture. As a sequel of former activities we also keep a low-noise maintenance activity of the PlanetLab software, which has been running in particular on the PlanetLab global testbed since 2004, with an ad-hoc federated model in place between PlanetLab Central (hosted by Princeton University) and PlanetLab Europe (hosted at Inria) since 2007.

  During 2012, we have focused on the maturation of the SFA specifications and the SfaWrap codebase, with several objectives in mind. Firstly, we have contributed within the GENI (N.S.F.) project to the specifications of the Version 3 of the AM-API (Aggregate Manager API), which defines the primitives that a testbed management infrastructure has to provide in order to be SFA-compliant.

  Secondly, knowing that our former SFA implementation was targeting PlanetLab testbeds only, we needed on the one hand, to make generic this SFA implementation, by completely redesign and refactor its codebase, and on the other hand, we needed to support all the resources allocation strategies supported by the testbeds, namely the allocation of both 'shared' and 'exclusive' resources. As a result of this redesign and development effort, out new SFA implementation is now disseminated and started to be known, under the name of SfaWrap, and we believe that it can be used as a production-grade alternative to quickly add SFA compatibility on top of many heterogenous testbed management frameworks.

  Finally, in order to allow the community of networking researchers to execute cross-testbed experiments, involving heterogeneous resources, Planète team has been instrumental in federating a set of well-known testbeds through the SfaWrap, namely PlanetLab Europe, Senslab - developed in other Inria Project-teams -, FEDERICA, the outcome of another E.U.-funded project and more recently NITOS, an OMF-enabled wireless testbed. See  [96] and  [97] for more details.

   

• Content Centric Networks Simulation

   

  We worked this year on the extension of the DCE framework for ns-3 in order to run CCN implementation under the ns-3 simulator. DCE stands for Direct Code Execution, its goal is to execute unmodified C/C++ binaries under ns-3 network simulator. With this tool researchers and developpers can use the same code to do simulation and real experiments. DCE operation principle is to catch the standard systems calls done by the real application in the experiment and to emulate them within the ns-3 virtual network topology. Concerning CCN we use the PARC implementation named CCNx which is a well working open source software reference implementation of Content Centric Network protocol. As promised by DCE this integration of CCNx requires no modification of its code, it requires 'only' working on adding the system calls used by CCN that are not already supported by DCE. The advantage of this approach is that the integration work of CCN advanced DCE and will be useful in others completly different experiments. Another great advantage is that every evolution of the CCNx implementation is very easy to integrate, all what is needed is to compile the new source code. The next steps will be naturally to use DCE/ns-3 to evaluation CCN protocols in specific scenarii, to improve the coverage of systems calls supported by DCE, and to improve the DCE scheduler to be more realistic and to take into account CPU time spent in router queues. This work is done in the context of the ANR CONNECT project and is currently under submission.

   

• ns-3 Module store

   

  Bake is an integration tool which is used by software developers to automate the reproducible build of a number of projects which depend on each other and which might be developed, and hosted by unrelated parties. This software is being developed with the participation of the Planète group and is intended to be the automatic building tool adopted by the ns-3 project.

  The client version of Bake is already working and the Planète group had a significant participation in its development. The contributions were in the context the addition of new functionalities, bug fixing and in the development of the regression tests. We are now starting the development of the ns-3 modules repository, which is a web portal to store the meta-information of the available modules. In the present state we have already designed and implemented the portal data basis and the main interface. It is already possible to register new modules and browse among the already registered ones.

  The web portal has to be finished, notably the part that will create the xml file that will be used to feed the bake's client. We also need to add new functionalities to the client part, to enable incremental build over partially deployed environments. As it is today, bake does not enable the user to add just one new module to an already deployed version of the ns-3 simulator. This work is done in the context of the ADT MobSim in collaboration with Hipercom and Swing Inria project-teams. For more details see the Bake web page http://planete.inria.fr/software/bake/index.html

   

• The ns-3 consortium

   

  We have founded last year a consortium between Inria and University of Washington. The goals of this consortium are to (1) provide a point of contact between industrial members and the ns-3 project, to enable them to provide suggestions and feedback about technical aspects, (2) guarantee maintenance of ns-3's core, organize public events in relation to ns-3, such as users' day and workshops and (3) provide a public face that is not directly a part of Inria or NSF by managing the http://www.nsnam.org web site.

   

• Automated Deployment and Customization of Routing Overlays Across Heterogeneous Experimentation Platforms

   

  During the last decades, many institutions and companies around the world have invested great effort into building new network experimentation platforms. These platforms range from simulators, to emulators and live testbeds, and provide very heterogeneous ways to access resources and to run experiments.

  Currently, a growing concern among platform owners is how to encourage researchers from different platform communities to take advantage of the resources they offer. However, one important aspect that needs to be overcome in order to appeal researchers to use as many experimentation platforms as necessary to best validate their results, is to decrease the inherent complexity to run experiments in different platforms. Even more so, to decrease the complexity of mixing resources from different platforms on a same experiment, to achieve the combination of resources best suited to the experiment needs.

  To address this concern, we developed the Network Experiment Programming Interface (NEPI) whose goal is to make easier the use of different experimentation platforms, and switch among them easily. The development of NEPI started in 2009 with the implementation of the core API, an address allocator, a routing table configurator, but also a prototype ns-3 backend driven by a simple graphical user interface based on QT. On 2010 we validated and evolved the core API with the addition of a new backend based on linux network namespace containers and stabilized the existing ns-3 backend.

  During 2011, we enhanced the design of NEPI and provided experiment validation, distributed experiment control, and failure recovery functionalities. In particular, we enforced separation between experiment design and execution stages, with off-line experiment validation. We also introduced a hierarchical distributed monitoring scheme to control experiment execution. We implemented a stateless message-based communication scheme, and added failure recovery mechanisms to improve robustness. Also on 2011, we started work on a prototype PlanetLab backend.

  Last year, we extended NEPI to provide automated deployment and customization of routing overlays using resources from heterogeneous experimentation platforms. The main contribution of this work is to enable researchers to easily integrate different resources, such as simulated, emulated or physical nodes, on a same experiment, using a network overlay, thus addressing one of the main concerns previously mentioned.

  We started by adding support to easily build routing overlays on PlanetLab, and providing the ability to customize network traffic by adding user defined filters to packets traversing the overlay tunnels [48] . We then improved this work by adding the ability to include simulated nodes from the ns-3 backend and emulated nodes from the linux containers backend into a single overlay network. We demonstrated the use of NEPI to build adn control routing overlays which incorporate resources from different on the ns-3 2012 community workshop [74] .

   

• Content Centric Networks Live Experimentation

   

  Realistic experimentation on top of Internet-like environments is key to evaluate the feasibility of world wide deployment of CCNx, and to assess the impact of existing Internet traffic conditions on CCN traffic. However, deploying live experiments on the Internet is a difficult and error prone task, specially when performed manually.

  To address this issue, during the last year, we extended NEPI, a framework for managing network experiments, to support easy design, and automated deployment and control, of CCNx experiments on the PlanetLab testbed. Among other features, NEPI now enables the deployment of user modified CCNx sources on arbitrary PlanetLab nodes, and the creation of tunnels to enable the use of multicast FIB entries between CCNx daemons over the Internet. By supporting easy CCNx experimentation on PlanetLab, NEPI can help to explore the co-existence of CCN and TCP/IP architecture.

  This work was presented as a poster and a demo at CCNxCon 2012, the CCNx http://www.ccnx.org/ community meeting [73] . The work had a very good reception and gained NEPI some new users.

  An online tutorial and demo were also made available at NEPI's web page http://nepi.inria.fr/wiki/nepi/CCNxOnPlanetLabEurope , for dissemination purposes.

   

• Smooth-transition: a new methodology for dealing with various network experiment environments

   

  The smooth-transition is a new methodology, which supports various network experiment environments covering from pure simulation through realistic emulation consistently. The reproducibility in experimental network research is getting important feature for iterative experiments in short-term and long-term period. The main idea of this concept is providing the reproducibility in a broader sense. So far, we had to implement different experiments by different environment, such as simulation, application-level emulation, and link-level emulation. Whereas the smooth-transition is able to keep the context of the experiments started from a pure simulation up to a realistic emulation gradually. That means the user does not need to waste time any more for learning and following a lot of documents and manuals from each different environment. Moreover, anyone can easily start to use the testbed and to develop inside (i.e. protocol stack). Because NS3 which is the most popular and powerful network simulator has been used in this concept as an experiment engine.

  The smooth-transition employees Network Experiment Programming Interface (NEPI) to conduct all functions, such as composing scenario, node deployment, experiment control, and resource management. The core of building this concept is NS3 which has Emulation (EMU) and Direct Code Execution (DCE) modules. EMU supports to use real network devices instead of NS3 MAC and PHY layer implementations. DCE is able to launch real application on top of NS3 protocol stacks. Furthermore, real Linux kernel (currently, net-next 2.6 is available) can replace NS3 Internet protocols by its advanced mode. This concept needs back-end system covering all experiment nodes. Control and Management Framework (OMF) plays an important role as a software framework to control and manage an wireless network testbed, and all messages are exchanged by Extensible Messaging and Presence Protocol (XMPP). Nitos scheduler has been adopted as a reservation system http://nitlab.inf.uth.gr/NITlab/index.php/scheduler . The user can reserve a time slot, nodes, and wireless channels through its web page. In addition, SFA supports that the testbed is federated with other ones of outside.

  The testbed provides PCAP files as a common outcome, and this file contains captured in and out packets. However, the file size is easily over gigabytes, then it makes a very long delay to process dozens of that files. To reduce the processing time efficiently, we are using an indexing scheme for fast collecting desired packets by filtering. In particular, this scheme is very useful to find packets occurred rarely, when an detailed analysis is required for an network event, such as retransmission, intrusion detection, and node association/disassociation. The indexing information is stored in a database file, and it does not need to be modified after making the file. The size of the file is very small compared with the PCAP file, so it provides fast packet filtering permanently, even after leaving the testbed. This work, post-processing of PCAP files, is in a collaboration with Diego Dujovne and Luciano Ahumada from the Universidad Diego Portales of Chili. Especially, YoungHwan Kim, a postdoc of the Planète group, has been currently dispatched for this collaboration for fourteen weeks (September 15 2012 $\sim$ January 26 2013) in Santiago, Chile.

   

• The FIT experimental platform

   

  We have started, since 2011, the procedure of building a new experimental platform at Sophia-Antipolis, in the context of the FIT Equipment of Excellence project. This platform has two main goals : the first one is to enable highly controllable experiments due to its anechoic environment. These experiments can be either hybrid-experiments (as NEPI will be deployed) or federated-experiments through several testbeds. The second goal is to make resource consuming experiments ( like CCNx ) possible due to some powerful servers that will be installed and connected to the PlanetLab testbed. During 2012, the specifications has been defined and the procedure will continue during the next year.

   

• Network Simulations on a Grid

   

  We studied an hybrid approach for the evaluation of networking protocols based on the ns-3 network simulator and a Grid testbed. We analyzed the performance of the approach using a simple use case. Our evaluation shows that the scalability of our approach is mainly limited by the processor speed and memory capacities of the simulation node. We showed that by exploiting the emulation capacity of ns-3, it is possible to map complex network scenarios on grid nodes. We also proposed a basic mapping algorithm to distribute a network scenario on several node [32] .