A Testbed of terabit IP Routers rUnning MPLS over DWDM

The results achieved by FUNDP during the ATRIUM project are of three types. First, FUNDP has developed new traffic engineering methods that can be used to control the flow of the interdomain traffic of Internet Service Providers (ISPs) or Corporate networks. Discussions with ISPs and National Research Networks (NRN) indicate that those networks often need a better control on their interdomain traffic, either for economical or performance reasons. The research on the development of those methods will be pursued by the ATRIUM researchers after the end of the project. A second type of result is that the researchers working on the ATRIUM project have contributed to the open source J-Sim simulation environment by developing models of MPLS and RSVP-TE, including the extensions proposed within IETF. The second software contribution is the development of a new an efficient simulator called C-BGP. This simulator is able to accurately reproduce the routing choices performed by BGP in large networks. These two open source packages are available for researchers outside the ATRIUM project. The third type of result is related to the educational mission of the University. The knowledge gained within the ATRIUM project has allowed us to improve the undergraduate networking courses and was used to provide new training courses for industry on advanced networking topics, including traffic engineering and BGP. The ATRIUM project also had an impact on the graduate-level education since three PhD. have been started within the project and one of that PhD is already almost finished. Finally, the ATRIUM results have allowed the researchers working on the project to propose new research projects to pursue the research activity.

Within the ATRIUM project, France Telecom has built a major part of a pan-European high speed optical IP core network interconnecting the IP research labs of Alcatel Belgium, France Telecom Paris, Telefonica I+D Madrid, PSNC Poland and the University of Liege. During the network construction and operation the experience gained allows to - Understand and satisfy the requirements of the high speed networking research community both from network and demanding applications point of view; - Assess the performances of a real sized network without impairments on real customers; - Cooperate with national or paneuropean research networks such as Renater or Geant. The cooperation with a European router manufacturer give a deep insight on architecture and features of what should be a carrier grade high end router compared to the basic opaque black box view provided by most manufacturers in regular RFI or procurement processes within France Telecom. The inter-domain and intra-domain traffic engineering algorithms proposed in WP3 and WP4 to establish Label Switched Path (LSPs) in an optimised fashion will be considered to improve the bandwidth utilisation and operation of the France Telecom network. The testing tools and methods enhanced and developed within the Atrium project framework allows to characterize high-end network elements for conformance, interoperability and performance both - In data plane at OC48 or higher bit rates; - In control plane by emulating complex networks of 1000s of routers. These tools are to be reused for performance and quality assessment in the framework of France Telecom procurement process, industrial partnership or collaborative projects.

Summary of results: We have developed DAMOTE (Decentralized Agent for MPLS Online Traffic Engineering), which provides two main basic functionalities: - QoS-based routing of DiffServ LSPs (Label Switched Paths) under constraints. - Local detour (backup) LSP routing for fast restoration. The first main function of DAMOTE is to compute primary paths at ingress nodes, in a way similar to the classical CSPF (Constraint Shortest Path First). This means that all edge nodes will compute and set up the best path for any given LSP for which they are the ingress. This computation requires that ingress nodes have enough information about all link states in the network. This is usually achieved by using extensions of link-state routing protocols like OSPF-TE or ISIS-TE, which flood the network regularly with updated link-states. Although similar in principle to CSPF, our scheme generalizes it in several ways. While CSPF is a simple SPF on a pruned topology, obtained by removing links that have not enough resources to accept the new LSP, DAMOTE can perform much clever optimisations based on a network-wide score function. Examples of such functions are: load balancing, hybrid load balancing (where long detours are penalized), pre-emption-aware routing (where LSP reroutings are penalized). DAMOTE is generic in the sense that this score function is a parameter of the algorithm. Like in CSPF, constraints can be taken into account, but here again the constraints can be parameterised quite freely. Typical constraints refer to the available bandwidth on links per class type (CT), or to pre-emption levels. DAMOTE can also compute local detour LSPs for fast rerouting. In our approach each primary can be protected by a series of detour LSPs, each of them originating at the node immediately upstream of any given link on the primary path. Those detour LSPs thus protect the downstream node (if possible) or the downstream link and merges with the primary LSP anywhere between the protected resource and the egress node (inclusive). Those many LSPs have to be pre-established for fast rerouting in case of failure, and provisioned with bandwidth resource. In terms of bandwidth consumption, this scheme is only viable if detour LSPs are allowed to share bandwidth among themselves or with primary LSPs, which is what we have achieved. The characteristics of these algorithms are: efficiency, near optimality, genericity, scalability, and pre-emption levels awareness. Prototyping and testing: Besides simulations, these algorithms have been implemented in ANSI C to achieve efficiency and portability. The result is a process that has been integrated in Linux and tested on a testbed of 6 Linux PCs configured as MPLS routers. The tested functionality includes primary LSP computation and setup with various score functions, and LSP hard and soft pre-emption and rerouting. The DAMOTE agent is ready for integration in any other platform. Link with IETF: Our solutions are aligned with the current work at IETF, and in particular with the Bandwidth Constraint (BC) models of the TEWG. In all cases we rely on the proposed TE extensions of OSPF and RSVP currently being standardized. For some algorithms we need a few additional objects in OSPF-TE and in RSVP-TE PATH and RESV as summarized below. Application domain: It should be noted that the overall philosophy has always been to find efficient distributed solutions suited for online traffic engineering. Our solutions are scalable and nearly optimal, despite the fact that the LSPs are established on demand and in sequence. Thanks to its performance, our solutions can thus support the automatic computation and establishment of MPLS-based VPNs at any timescale. In a hierarchical network where inter-domain LSPs are established, our algorithms can also be used to compute and set up the intra-domain parts of those LSPs. It is also worth noting that our tools are targeted at strategic TE rather than tactical TE. In tactical TE, LSPs are set up and removed on-demand to circumvent some temporary congested areas, based on some real-time monitoring of the network. In strategic TE, the LSPs are set up to guarantee a certain level of QoS a priori. However, given the very low computation time of an LSP (usually less than 100ms, even when several class types and pre-emption levels are considered, which remains less than the LSP setup itself), our solutions could also be used for tactical TE. It is even more so because the computation algorithm can take into account the congested links that should thus be avoided, by relying on the existing mechanism that already allows us to compute a new path for a rerouted LSP without reusing the link on which it was pre-empted. Dissemination: Our work has lead to 8 papers among which 6 are already published. Our results have been presented at 5 conferences and/or workshops.

The result of the tasks is to compare the quality-of-service (QoS) possibilities of the A7770 Xantium router with an existing QoS implementation on PC based routers (based on the MIT Click modular router toolkit) and with a Cisco 7507 platform. This comparison would be both in functionality and performance, especially from a DiffServ viewpoint. The functionality analysis includes testing and comparing the routers as DiffServ Edge, Leaf or core router as far as they support the different functions as e.g. marking, shaping, policing. Also functionality as per-hop-behaviour (PHB) configuration (e.g. for the standard EF and EF DiffServ classes) will be studied. The performance testing was done by using the Spirentcom Smartbits platform with Terametrics modules and a.o. the Teracaw software. This software is based on the use of a real network (TCP) stack and can be used to emulate a lot of FTP/HTTP users and servers e.g.. The real TCP stack is especially needed to measure the impact of RED (Random Early Detection). Besides these tests, the normal packet performance measurements (a.o. with SmartFlow) will also be done to test the different behaviours (delay, jitter) of different QoS classes. The PC based routers are used as a kind of reference for the functionality comparison (of course, the performance comparison will be a relative comparison) as we have developed in-house a CORBA interfacing on top of the MIT modular router toolkit, which implements the DiffServ MIB (RFC 3289, converted according the JIDM SNMP-CORBA conversion standard).

The ATRIUM project has built a pan-European high-speed optical IP core network interconnecting the IP research labs of Alcatel Belgium, France Telecom Paris, Telefonica I+D Madrid, PSNC Poland and the University of Liege. The way the network was built and operated is for an equipment vender an experience and allows understanding the difficulties of the customers of Alcatel in operating and maintaining a commercial IP core backbone. ATRIUM designed and implemented Layer 2 virtual private networks using MPLS and advised NRENs and Dante how to provision such a networks on their IP core backbone. During the project execution, the ATRIUM project presented its results at 37 conferences and two-demonstration exhibition, wrote 28 papers and delivered 11 IETF drafts, organised 16 meetings with NRENS, IST projects and GEANT. This project introduced Alcatel in the Optical IP research community of Europe and increased the credibility of the company as an IP player in the international forum. The interoperability test result determine that the testbed built can be put in real operational conditions facing heterogeneous network elements at intra- or inter-domain level. The results obtained during the reporting period show that the different Alcatel IP core-A7770 releases are capable to support QOS, MPLS traffic engineering and multicast scheme. The European manufactured IP core router can be released to the European Research companies and to the network operators. The inter-domain and intra domain traffic engineering propose and asses protocols and algorithms for intra-domain traffic engineering in MPLS networks and designs on-line decentralized traffic engineering algorithms to establish Label Switched Path (LSPs) in an optimised fashion. These algorithms can be implemented in the IP portfolio of Alcatel and will improve the bandwidth utilisation and operation of the network deployed with Alcatel IP products.

The results achieved during ATRIUM project are focused on the use of demanding applications on a terabit pan-European network with Quality of Service in IP/MPLS based infrastructure and on high-speed network performance evaluation. PSNC also benefits from the experience with the new technologies for high-speed optical networking and large-scale service provisioning that will support building of the Polish Optical Internet. PSNC is a high performance computing center and therefore uses GRID-related applications on a day-to-day basis. The ATRIUM project gave us an experience in running these types of applications in a large, high-speed network. The results will be exploited in the Polish Optical Internet PIONIER, with its 10G backbone connecting all metropolitan area networks in Poland and 5 high performance computing centers. Running large file transfers or distributed computations over ATRIUM testbed, PSNC had a possibility to verify the design of terabit optical network and the adequacy of the proposed solutions with respect to the requirements of very demanding users. The work conducted during the course of the project will support the development of GRID-related services and future GRID-related projects especially when collaborating with remote sites within Europe. The results and QoS techniques assessment will ease the initial configuration of GRIDs and advanced bandwidth consuming applications. The results of the project will be used not only within national environment but also adopted to a European scale and used for international collaboration e.g. during FP6. As the manager for Polish NREN, PSNC is responsible for introduction of new networking technologies and the network expansion. Currently the main objective of PSNC is to complete building of the optical network connecting all major universities towns in Poland � PIONIER. This network is based on more than 5000km of owned optical fibers which (supported by accurate technology) can provide enough capacity for research community in Poland. The IP/MPLS ATRIUM network architecture achievements together with state-of-the-art optical DWDM technology will strongly support building of the Polish Optical Internet in its next phase of optical development. Quality of Service and Traffic Engineering mechanisms designed and implemented in ATRIUM testbed will also be used during the network development. Since PSNC was connected to the ATRIUM testbed with the use of (for the first time) CCC (Circuit Cross Connect) connection through the GEANT network it gave us a possibility to exploit the advantages of that new technique for interconnecting project remote sites. That experience helped us to deploy new connection for another FP5 project as well as will save unnecessary operational work for future projects. The main end-users of the results will be researchers especially from the GRID community. They include researchers from PSNC as well as Polish research community gathered in the current POL34/622 network, which constantly runs GRID applications and GRID-related projects (either national or international). The results also benefit operation department of Polish NREN that develops PIONIER network and introduces new services.

The main results regarding Interoperability testing are: - The experience gained in the integration and testing of a Terabit IP Core Router (ALCATEL 7770 RCP) in large IP/WDM networks with MPLS, verifying the interoperability issues and gaining experience in the integration of this kind of equipment in IP networks. - Experience in multicast PIM-SM, testing and verifying the main procedures related with multicasting in large IP networks, as RP's selection, trees instantiation and reliability. The main results regarding Testbed Experimentation and Demonstration are: - Experience on testing and integration of advanced multimedia services on large IP networks. - Experience on configuring and testing QoS parameters in IP-MPLS networks, based on DiffServ specifications and verifying the interoperation with multicasting procedures. - Knowledge of the main issues and strategic decisions in which the deployment of QoS and multicast policies are based. - Experience on definition and deployment of a QoS/multicast scheme over a pan-European IP/WDM MPLS network, mapping and applying the scheme with advanced multimedia services deployed in the network. The main result regarding network management is the availability of an OSS prototype (named AtriOSS) with the following main characteristics: - Network performance management. It has a twofold goal. On the one hand, it aims to provide the proper tools to identify what quality related problems are affecting the network (essentially caused by congestion), where are they located and which is their impact on the carried traffic. Two sources of information will feed AtriOSS network performance databases. Local measurements taken by routers and end-to-end active measurements collected by distributed monitoring stations - Service quality management. Post-processing of raw measurements allows estimating the actual quality delivered to customers. This is compared against SLA defined parameters; the contents of each signed SLA is represented by AtriOSS in the form of a structured, formal Service Level Specification (SLS). An SLS instance contains the quality parameters under agreement, the objective values for each parameter, penalty clauses if applicable and any other information that may affect the service (for instance, considerations on planned outages treatment). - Network inventory. AtriOSS will provide a full view of the network under management. It will keep an inventory of both the physical (nodes, racks, cards, etc�) and the logical (LSPs, tunnels, subnetworks, meters, classifiers, etc.) network entities. - Service provisioning. AtriOSS addresses some of the functions traditionally associated with this management layer as support for its SLA management capabilities. AtriOSS differentiates four phases in the provisioning process: service definition, service deployment, service subscription and service usage. - Fault management. AtriOSS supports the collection, synchronisation, filtering, logging and presentation of alarms coming as SNMP notifications from network nodes. Alarms are modelled using X.733 [12] format.

Deliverables

Share this page

Download