Laboratories over Next Generation Networks

The aim of this document is to draw the conclusions and results obtained in the LONG project by September 2002 (M22) as requested in the first project review (Apr 5th 2002, Madrid). This document was requested as a draft version since the final one will contain all the summary of conclusions and results obtained until the end of the project in January 2003 (M26).

This report states the planned activities to be carried out by the project LONG within the work package 5 about dissemination and implementation. This document was delivered at month 6 since the beginning of the project. It includes the first dissemination activities performed during this period of time and a plan of the ongoing and planned activities for the rest of the duration of the project.

Netperf is a benchmark that can be used to measure various aspects of networking performances. Its primary focus is on bulk data transfer and request/response performance using either TCP or UDP and the Berkeley Sockets Interface. Netperf IPv4 is owned by Hewlett-Packard. Netperf for FreeBSD has been modified to run tests with SOCKS64 and NAT-PT(4->6) within the LONG project.

The aim of this document is to draw the conclusions and results obtained in the LONG as requested in the first project review (Apr 5th 2002, Madrid). This document presents technical conclusion obtained during the whole project regarding network design, deployment and testing, user services and application migration.

The aim of this document is to review the impact on collaborative applications over IPv6 when specialised components are used. The document has been completed with descriptions of some migration experiences. The applications selected are classified as simple multimedia distribution tools, such as the streaming video tool, or collaborative tools such as some multi-user distributed games. This document is the second phase of work summarized in Deliverable D3.2.

This document describes the network infrastructure deployed by the LONG project. Several aspects are referred to network topology, addressing and routing, basic and advanced network level services and, adopted transition mechanisms, just to mention some. The network described in this document was thought to make possible the execution of trials using IP services and applications on IPv4/IPv6 mixed scenarios.

MAngband is a free online multiplayer real-time roguelike game, derived from the single player game Angband (computer role playing game based loosely on the world of J.R.R.Tolkien, in which the player explores the complex dungeon of Angband, killing monsters and collecting treasure, and eventually defeating Morgoth, the Master of the Pits of Angband). To win the game you must build up a character, selected from any one of 10 races and 6 classes, to be powerful enough to defeat Morgoth, Lord of Darkness, who lives in the dungeon 5000 feet below the surface of the town. As you work towards this near-impossible goal, you will chart unexplored dungeon passageways, do battle with legendary creatures, and find lost artifacts. The work made in LONG has consisted on migrating this free version of MAngband working over IPv4 to IPv6.

This document reports a description of the publications, contributions and dissemination events of the Project. It is, thus, a compilation of the dissemination activities of the project. It includes papers published in journals, papers presented at conferences, RFC submitted to IETF, presentations done in several international events, presentations of the activities of the project given in meetings of other projects, guidelines published, list of attendance to co-ordination activities with other projects including IPv6 Cluster meetings, demonstrations organized directly by the project and demonstrations where LONG participated. All these activities are well documented in the website of the project: http://www.ist-long.com.

Tetris game has been migrated to IPv6 in the LONG project. The distributed game based on the client-server paradigm has been developed outside LONG, the work made in LONG consisted on migrating both, the server and the client, to IPv6. To migrate Tetris patches have been used, so that to have an IPv6 capable game, you have first, to install the free software of Tetris and then apply the IPv6 patch. For Tetris server "tetrinex" has been used (tetrinetx-1.13.16+qirc-1.40c). While for client, "gtetrinet" has been taken as basis (gtetrinet-0.4.1).

This Report shows the results concerning trials whose specifications were described in LONG Deliverable 4.1. The mechanisms tested are described in Deliverable 2.1. The methodology used and the tests performed have been previously presented in the Deliverable 2.1 of the LONG project. The tools required, MGEN and Netperf, have been adapted specifically for this purpose, and are available in the web of the project. The tests corresponding to the first phase are carried out in a local basis.

This is an example of a game developed in IPv4, outside the project, that has been migrated to IPv6, inside the project. Chess is a wide-known game that is played through the networks based on the client-server paradigm. The server used in the LONG network was IPv6 only and the client IPv4/IPv6, thus access to the LONG Chess server was allowed, through IPv6, and the client was also able to access any other IPv4 server in the network.

This Deliverable defines trials and experiments to obtain feedback on network performance and functionality. These trials will be focused on testing transition mechanisms, network access mechanisms and advanced services. The conclusions obtained will ease the use of the collaborative work application that will be developed in WP3 over the network deployed in WP2.

In order to provide QoS support for IP networks, a software application must integrate all (or most of) the measurement and processing tools related with the experiments into a single environment. This application provides an integrated graphical interface for a set of tools that allows the measurement of QoS parameters over IPv4 and IPv6 networks. Network Performance Meter (netmeter) is an application, which tries to solve these tasks in a flexible and modular way.

MGEN is a set of programs, which provide the ability to perform IP network performance measures using UDP/IP unicast and multicast traffic. The toolset generates real-time traffic patterns so that the network can be loaded in a variety of ways. Script files are used to drive the loading patterns over the course of time. It is possible to create script files, which emulate the traffic patterns of unicast or multicast UDP/IP applications. The receive portion of this tool set can dynamically join and leave IP multicast groups and log receive data and can be used to calculate statistics on received data such as throughput, packet loss rates, communication delay, etc. RSVP operation in conjunction with ISI's "rsvpd" Release 4.1aX is supported and RSVP reservations can be dynamically scripted as well. UC3M has ported Mgen to IPv6 for the LONG project.

LONG network statistics generation scripts consists of a tool that checks every configured time the losses and delays from a site to a set of nodes and paints a jpeg file with the result, so that they can be showed in a web page, for example. In LONG thanks to these scripts the status of the LONG network is accessible through the web page, showing the status of all the routers used to interconnect the partners, characterized by the losses and the delays.

This deliverable reports the results from several experiences over the advanced IPv6 environment and provides some insights valid for IPv6 deployment. It presents the status of the network and the results of link availability and performance tests. Link performance tests have been identified as a key test to provide information about misconfigurations that cannot be easily discovered otherwise. Distributed tests involving more than one partner are also presented: DNS (if it can be considered as a network service, due to its core importance for the deployment of other applications), multicast, security, mobility, anycast or multihoming. Distributed applications have also been tested. Among the IPv6 user applications experimented, we find Isabel, HTTP, FTP, IRC, News, Mail and LDAP. Most applications have been successfully deployed over a mixed scenario (including the complex Isabel application). Some of the local tests (DNS and HTTP) are focused on performance. QoS tests have also been thoroughly executed locally. Other local tests made are multicast or mobility, anycast or NFS (that is by nature a service to be provided on a local scope). There is also a report on the experiences involving real users: Isabel and the rest of the applications have been available in many public events for real users. Among these events, we highlight the IPv6 Forum 2001 and 2002 held at Madrid, the Campus Party held at Valencia, and the IST 2002 Event held at Copenhagen.

The aim of this tool is to allow IPv4 only hosts to make ping to hosts that support 6to4 tunnels. To develop this tool the author has used the ping implementation described in "UNIX Network Programming - Networking APIs: Sockets and XTI" by R. Stevens.

This document is a summary of migration guidelines described in detail in previous deliverables D3.1, D3.2 and D3.3. The document is divided in three parts. The first analyzes existing applications to look for characteristics, which usually should be reviewed during migration process. The second is devoted to show the basic socket interface extensions for IPv6, fully described in RFC2553. Finally the last part provides general recommendations for new IPv6 applications. This document provides guidelines to migrate applications to new IPv6 environment.

This document provides guidelines for migrating applications and services inside distributed laboratories. The porting section includes the evaluation of strategies, the set up of dual-stack communication framework and the adoption of new communication libraries. The document provides evaluation mechanisms and review of some concrete application porting examples.

LONG IPv6 web tools is a set of network tools widely used in IPv4 and very useful that has been made accessible through the web page. These tools allow all the IPv4/IPv6 users that have access to http, use them from an IPv6 host within the LONG network to the IPv6 networks where LONG has access, including 6Bone. The tools include: ping, mtr, traceroute, nslookup, reverse nslookup, mail, LDAP queries (to an LDAP server available in LONG). Besides the tools commented before, LONG IPv6 WEB TOOLS also include other useful and more advanced tools such as statistics and irc and includes information of different type of access using NAT-PT or proxy.

This document describes the study performed in the LONG project for the deployment of network and application level services in mixed IPv4/IPv6 scenarios. The theoretical and functional aspects of services are presented. Also, this document identifies the services that are being deployed on LONG infrastructure and how those will be used in order to evaluate their behaviour in a near-real environment.

This document describes a complete set of IPv6 implementations over different access technologies. Also, the functionalities of the available IPv6 equipment used in the trials and experiments achieved in WP4 are presented. As a result of the experience obtained, the access technologies that will be implemented in LONG project are described.

This document reviews the main characteristics of distributed collaborative applications and provides guidelines to migrate them to new IPv6 environment. The first part is devoted to review collaborative application characteristics from the communication point of view. It includes some general migration recommendations because detailed recommendations are not possible because of application architecture dependencies. The second part is devoted to detail the migration process of ISABEL application.

This Deliverable defines trials and experiments to obtain feedback on network performance and functionality. These trials will be focused on testing transition mechanisms, network access mechanisms and advanced services. The conclusions obtained will ease the use of the collaborative work application that will be developed in WP3 over the network deployed in WP2.

This report presents a brief description of the structure and contents of the web sites of the project, both, the internal one and the public one. As the use of the private web is mainly devoted to the partners that have been involved in the project and the public one is available for the general public, a more detailed description of the public one is presented.

The ISABEL CSCW application is a group communication tool for the Internet, based on advanced videoconferencing features, allowing efficient organisation of working procedures over the Internet in large enterprises or groups, such as - Minimization of travel or movement of persons; - Timely organization of meetings when needed; - Increase in the overall productivity of an organization; - Reduction of project and activity lead times. ISABEL uses an innovative service concept, which adapts the collaboration sessions to the user needs. It includes service definitions for different usage scenarios. Each service will support the specific behaviours and characteristics of a given set of users. For example a service for distributed meetings, for distributed lectures, for distributed congresses, etc. Isabel platforms are constructed over standard IP networks using either IP unicast, IP multicast or mixtures of both. A variety of underlying networking technologies can be used to construct the platforms, which may include Internet, GEANT, NRENs, Internet 2, Canarie, LANs, etc. ISABEL is a well proven technology, which has been extensively used and validated in real and world wide service trials and has several distinguishing features over standard videoconferencing: - ISABEL makes multi-point collaboration easy, effective and manageable with large numbers of endpoints. Effective collaborations with up to 20 sites have been performed; - ISABEL collaboration platforms are cheap, as no MCU is needed to set up multi-points and all the processing is just done by software. No special hardware is needed, just plain PCs; - ISABEL is based on TCP-UDP/IP and works over variety of networks or technologies without gateways, such as Internet, corporate networks, ATM, ISDN, Frame-relay, satellite; - ISABEL creates a new service concept, which has proven very effective in: corporate tele-meeting or tele-training; distance learning; or distributed congresses.

This Deliverable defines trials and experiments to obtain feedback on network performance and functionality. These trials will be focused on testing transition mechanisms, network access mechanisms and advanced services. The conclusions obtained will ease the use of the collaborative work application that will be developed in WP3 over the network deployed in WP2.

Livrables

Partager cette page Partager cette page sur les réseaux sociaux

Télécharger Télécharger le contenu de la page