IPv6 QoS Measurement

This document describes the approach undertaken in the project 6QM to the dissemination of project results into the target communities, and lists activities for creating broad public visibility to project results. The Dissemination Plan is detailing the opportunities planned to be taken by the 6QM consortium, and each partner's plan to exploit project results. The project consortium anticipates that wider application of our results will provide key functionality to support for a fast transition towards IPv6 infrastructures. Given the impact and relevance of the issues addressed by the 6QM project, the consortium considers the task of dissemination as highly important. In order to ensure that all information reaches its intended audience, the project has planned for a comprehensive communication strategy that addresses the particular information needs of each of its target customer groups. The project will produce material and documentation tailored to their specific information requirements. This input material is forwarded in the appropriate information channels and to industrial fora to ensure broadest diffusion. In addition to producing information content and making it accessible through the project website, 6QM adopts an pro-active role in disseminating its outcomes by participating in the organization of events and through presentations at conference, workshops and coordination meetings where project results will be discussed with a wider audience.

The main security issue in QoS measurement architecture is to secure signalling between all the actors permitting to collect data. Such architecture cannot work and provide concrete results to the ISP if its elements are incontrollable and give back bad data. The second important point is to ensure that data collected are the good ones (authentication of the sender, integrity of data and anti-replay) in order to set-up QoS-based services. Similarly, confidentiality of such measurements must be guaranteed in order to protect business. Customers may disagree with having the traffic of their applications to be analysed and so to be metered. Privacy may be important for customers so ISP have to respect it. Denial of Service (DoS) is not the main security issue but may be an important danger for an ISP with the loss of the quality of the services provided. Customers cannot accept to have their link to Internet disturbed or their IP based services disabled due to an attack on the provider networks. Thus, the detection and the suppression of DoS is necessary for an ISP. The information collected by QoS measurement systems can be used to detect DoS attacks. In this deliverable we devote one chapter to this issue. Measurements of the performance of IP based services require high tech systems. These systems should not be used to destroy the network they are measuring. This document illustrates all these points.

This first WP3 deliverable aims at providing the specifications of the 6QM Measurement System Prototype. In order to create these specifications WP3 tried to reuse as much as possible the concepts and guidelines developed in the WP2. As defined in the project technical annex the target is to develop a prototype measurement system capturing IPv6 packets in order to perform QoS measurements. Consequently the scope of WP3 is to address QoS measurements by means of passive measurement techniques. However as shown in WP2 active measurement techniques are very important for network operators. As a consequence, even if this technology is not in the original scope of the 6QM project, the current prototype specification addresses the possibility to include some active probing in the 6QM measurement infrastructure as a scope extension. Before specifying the prototype, this deliverable gives an overview of some relevant existing QoS measurement systems and compares them. This comparison points out that the active measurement is traditionally covered by many existing systems unlike the passive measurement. Moreover this study shows a need for passive QoS measurement systems in multi-point mode and a need for systems combining both passive and active techniques. This deliverable also fixes the 6QM measurement system scope by defining several levels of priorities for the prototype functions - referred as "Core Scope", "Extended Scope", "Advanced Scope" and "Future development". The definition of those boundaries addresses a strong need to clarify the prototype direction and to help in assigning the future development resource. Once this background information is clarified, the main part of the document presents the prototype specification by defining the initial metrics to be used in the prototype and by providing the system overview. The result of the system overview is to provide the basic system components defined as "6QM Measurement Manager", "6QM Evaluator" for the management and respectively data collection component. This system overview also defines some generic meter components, which could be passive or active (optionally). Then the deliverable describes the mandatory components in more detail, their external interface and their internal structure. Finally the document shows the relation between WP3 work and WP2 requirements. The present document defines the baseline for the prototype development. As far as the WP3 is concerned the next steps are the definition of the prototype detailed design and the development itself. The next deliverable (D3.2) will extensively address those issues.

The public trials are intended to define visible demonstrations to make the project results publicly available. Getting the result of the previous internal trial, in order to get the best combination of components and refine the testing scenario in order to show a proper interaction between all the project elements. The main target of such trials is to get a feedback from the field about the results and working of the prototype system. On the other hand, evaluation is also a critical activity to be performed in any project, as 6QM, dealing with new network architectures, adapted devices and next generation services and applications. The evaluation on 6QM project deals with the results gotten during the trials that have been performed during the project. Based on both the results and the feedback from the trials performed during the last part of the project, the OpenIMP system developed by the 6QM project is evaluated and the conclusions are presented in this deliverable, which in general validates the work done by the developers and also suggests some things that might be changed or improved.

This document examines the various working groups in the IETF that are relevant to the 6QM project. In particular, a general description of each working group is given, plus the goals and milestones, and the drafts and request for comments that are associated with each group. The following groups have been included: - IP Version 6 (ipv6): Provides a home for IPv6 work that spans multiple working groups. - IP Version 6 Operations (v6ops): Develops guidelines for the operation of a shared IPv4/IPv6 Internet and provides guidance for network operators on how to deploy IPv6 into existing IPv4-only networks, as well as into new network installations. - Benchmarking Methodology (bmwg): Focus is to make a series of recommendations concerning the measurement of the performance characteristics of various internetworking technologies. - IP Performance Metrics (ippm): Develop a set of standard metrics that can be applied to the quality, performance, and reliability of Internet data delivery services. - IP Flow Information Export (ipfix): Define a standard set of capabilities by which IP flow information can be transferred. - Packet sampling (psamp): Define a standard set of capabilities to sample subsets of packets. - Remote Network Monitoring (rmon): Chartered to define a set of managed objects for remote monitoring of networks. - Internet Traffic Engineering (tewg): Traffic Engineering entails that aspect of network engineering which is concerned with the design, provisioning, and tuning of operational internet networks. - Audio and Video Transport (avt): The Audio/Video Transport Working Group was formed to specify a protocol for real-time transmission of audio and video over UDP and IP multicast. - Inter-domain routing (idr): The objective is to promote the use of BGP-4 to support IP version 4 and IP version 6. The working group will continue to work on improving the scalability of BGP. Various standardized measurement architectures are described and then compared. These include, but are not limited to: - RMON: Remote Monitoring is a standard monitoring specification that enables various network monitors and console systems to exchange network-monitoring data. - IPPM: defines a MIB for managing the measures using the IP performance metrics specified by the IPPM Working Group. - RTFM: The RTFM architecture is an attempt by IETF to standardize several aspects of flow definition, capture and metering operations [RFC2722]. - Sflow: a technology for monitoring traffic in data networks containing switches and routers. - IPFIX: Currently defining an architecture that employs the concept of collector, observation point, and metering process.

ITU-T and IETF collaboration must be strengthened in the new industry emphasis on Internet and IP structured signals. Neither the IETF nor the ITU-T is able to adequately address IP based networks independently. For example, the IETF strength lies in the protocol and application areas, whereas the ITU-T has a great deal to offer in the areas of architectural, network interworking and network evolution. This document details the state of the art of the standardization of IPv4 and IPv6 at the ITU-T. In particular, it emphasizes the necessity of developing a standard test packet structure that can be used for both IPv4 and IPv6. In addition, the related activity will be in charge of the dissemination inside ITU-T recommendations of the work that will be realized by the 6QM project.

This document is a summary of the 6QM project more relevant information, which the main goal of developing a comprehensive approach towards IPv6 QoS measurement. Include the objectives, technical approach, key issues, and expected impact. Includes also the list of participants.

This document describes the work done in the initial public trials successfully done by 6QM during the first year in IST 2002, November 2002, CeBIT, February 2003, and Madrid 2003 Global IPv6 Summit, May 2003. In addition, considering the project extension, and because its relevance, the description of the most complete public demonstration of 6QM prototype showed until now during 6NET Spring 2004 Conference & Eurov6 Showcase, 18th-19th May 2004 in Brussels, is also included. The main goal with the initial Public trials was to show and test early demonstrators for the system under construction that allows outsiders to get a glimpse on the development work. Also, the idea is reach a good dissemination of project results and to facilitate cooperation with other projects.

The goal of this deliverable is to define the management architecture for QoS measurement such that users and service providers can have a common understanding of the performance and reliability provided by the Internet clouds that are traversed. Defining the management architecture involves the following: - Looking at existing passive and active measurement architectures and potentially extending and/or enhancing and possibly correlating them. - Examining the requirements for intra domain measurements. - Defining methodologies for intra-domain and inter-domain measurements. - Standardizing the exchange of measurement results among heterogeneous measurement systems and across administrative domains, thus allowing for concatenation of global metrics. - Defining the set-up process for creating end-to-end measurements across administrative domains. - Standardizing the format and semantics of test packets for interoperability between probes. - The Definition of a Management Level System Architecture - The examination of aspects related to mobility, security, and IPv6.

The project web site (www.6qm.org) has been established for on-line information exchange. This document describes its general outline and structure, the rationale and the type of information that has already been made available and what will be published as the project further progresses. All public information will be made accessible through this site. The web site has both IPv6 and IPv4 capability, so that the IPv6 connectivity can be demonstrated across several IPv6 networks. The site serves as repository for presentations and publications of the project partners. It acts as a public information service on topics related to IPv6 QoS Measurement offering links to other relevant projects and initiatives worldwide. The portal also holds a password protected private workspace for the 6QM project partners for on-line information management.

This document is an introduction to the study of the measurement of IPv6 QoS in home environments. It presents the heterogeneity of home networking environment from differences perspectives such as media, devices, applications and actors. In a way to propose guidelines for IPv6 QoS measurement in home environment it analyses the impact of home networking on privacy and analyses the impact on home environment on measurement.

Metrics are the basic elements that permit to have information from somewhere about something. In order to get relevant information, metrics are specific, dedicated to a special task, for example a round trip time or the time taken to download a file. Some of them can be used as it is, but sometimes we would like to have more information because: - There is a lack of measurement somewhere in the network. - The metric is not detailed enough (for instance the download time of a file is abnormally high but we don't know if it is due to the network or the application server). - The metric used does not look like the traffic we want to measure (for example, getting a round-trip time with a PING when we want to qualify a VoIP flow). The consolidation is a response to these issues: - Get the understanding of an (abnormal) event when not all the information is available. - Get the relevant information extracted from numerous metrics collected.

New challenges for network operators are being raised nowadays: new services, best cost-effective deployments, the IPv6 deployment, etc. All of them are on the way to let the network operator gain competitiveness in order to catch market share. Operators and ISPs must react quickly to these new demands of services with different profiles and with differentiated QoS. They must know and follow the behaviour of their networks in terms of traffic engineering, resource management, system management, network management, service management and business management. This is the global scenario of management, which is the aim, and the general scope of 6QM project. Besides this, the operators and ISPs need to optimise architectures, exploitation, engineering and dimensioning of their IPv6 networks and services with minimum costs. The results of IPv6 QoS measurement can be a good support for this optimisation, but it is only an immediate instantiation and implementation of the global management scenario mentioned before. Taking into account this minimum and early implementation of the management scenario, the operators and ISPs must be capable to control the QoS, to measure it objectively and to share information with the clients according the SLAs contracts. These measures are useful also to compare ISPs (comparison of SLAs provided by different ISPs). In these conditions, we must put on the hands of the operators and ISPs practical tools and operational rules. The former is already done by the 6QM project, which has been developed an IPv6-capable QoS monitoring system. The later has to be done at least by each ISP although it is highly recommended to setup minimal common practices carried out by all the service providers that help the network monitoring systems to be used on inter-domain environments. As part of the research done in 6QM project dealt with the issues derived from the measurements performed in large-scale networks, which can be directly applied to reach such an objective. In this document some guidelines for network monitoring with 6QM are presented. The objective is to illustrate briefly somehow what the 6QM system is, what can it does for service providers, what they need to deploy it and what considerations have to be taken into account. Through the diverse sections of this deliverable all these questions are faced to present a complete view of the utility of the 6QM system and to encourage ISPs and service provider in general its usage.

During the 6QM project many aspects of the IPv6 QoS measurement have been addressed. This document proposes a revision of the initial specifications presented in the deliverable D3.1 "IPv6 QoS Measurement Specification". The purpose of this document is to present some areas that could be improved or extended and specifies the means or techniques necessary to perform such an improvement. The investigated topics are presented hereafter. During the revised specification activity, important efforts have been put to enhance the inter-domain design that was introduced in D3.1. The design details the interface for set-up and the data export format between inter-domain agents. This document also makes proposals on how to reach security for the 6QM measurement system. The addressed topics are the secured communication between system components, and the restriction for accessing system components and user management. An important effort concerned the improvement of the system scalability. This examination includes techniques to increase performance for the network probes with the usage of sampling; but it also contains consideration on how to increase overall system performance by establishing extended levels of component hierarchies and a decentralized architecture. Finally a paragraph inspecting methods for measurement result storage concludes the argumentation about scalability. The document at hand follows the announcement given in the previous deliverable D3.1 to look for additional metrics of interest. A section presents metrics that we selected for further analysis including the metric for evaluating the available bandwidth. In a separate section, this document also provides some information about some prototype enhancements that have been performed after the original development time. Additional considerations that establish similarities and dissimilarities between existing tools that are listed by the MOME coordination action of IST and the measurement platform developed within the project frame of 6QM are given in a chapter of this document. The document concludes that there is still space for further improvement in design and implementation as introduced in the deliverable D3.4 addressing the guidelines for further research.

The 6QM Consortium identified several important kinds of tests to be performed on its prototype. Each one is needed to know different aspects of the accuracy on the most important parameters of the OpenIMP system, mainly the packet loss and precision of the measured one-way-delay. The evaluation of those two parameters is enough because all the calculations made by the OpenIMP system are based on them, so no extra tests are required. The tests have been performed on laboratory not only to know the reliability of such parameter but also, and maybe the most important, how different aspects like type of CPU, RAM amount on the system, number of measurement tasks running on the probes, etc. can influence on precision of the obtained results. As secondary objective, it was stated interesting to find out the limit traffic rate that the each architecture can support with reliable information. Regarding this point, it has been proved that 6QM probes installed on PIV 2.8 GHz are enough to capture packet on OC3 interface in a reliable way. The description of the tests, result and conclusions are presented on this deliverable.

This deliverable is the closing deliverable of WP3. During the 6QM project many aspects of the IPv6 QoS measurement have been addressed. The purpose of this document is to identify the areas where further research could be performed in the future. The document classifies the further research aspects into two categories: - Guidelines for further system improvement. - Guidelines for further application area. Concerning the system improvement the document identifies the following area: - Given the increase of traffic flowing in the network, the system should be able to deal with this increasing amount of data. This could be addressed by developing a hardware-based meter. - The inter-domain measurement remains a challenging issue and there is not yet some agreed standard to enable such a measurement. This topic still needs a great amount of work in order to deal with the existing heterogeneous systems and to create some technology that could be applicable by carriers at large scale. - The reliability of the system could be increased too. - The sampling should be investigated further. - This current measurement system addresses IPv4 and IPv6 however it could be interested to look forward at the new emerging address families. - The mobility was not addressed within the project and could be investigated in the future. The document identifies the related issues. - The control of heterogeneous systems should also be investigated in the future. Concerning the further application area the document identifies the following area: - The usage of the QoS monitoring to support autonomic communication. - The usage of the QoS monitoring to support the multi-homing in regard of the access selection. - The usage of the QoS monitoring to assist automatic MPLS path configuration. - The usage of the monitoring at application level would also bring value and this would need further investigation to target specific applications. Many areas remains opened in the field of the QoS monitoring. However one very important aspect that is pointed out in this document is the possibility to integrate core measurement components into existing system such as control and configuration systems. This integration could really bring benefit to the community.

The 6QM project (IST- 2001-37611) started on the 1st of September 2002 and initially was due to finish on the 31st of August 2004. However, after a request from the project team, the original work plan was extended for four months until the 31st December 2004. The reviewers deemed necessary a second extension, and the project was extended an additional four months until de 30th April 2005. The project goal was the research, development and integration of the different pieces needed for the correct measurement of the Quality of Service in IPv6 networks. In order to achieve this goal, a measurement device has been developed. This device inserts precise timestamp information when IPv6 packets are captured. A measurement server was also developed. This server collects the packets that were time-stamped by the measurement device and provide usage data and QoS metrics (delay, loss, jitter and so on). A demo was also developed, in order to show the potentiality of our system. It is worth to mention that our framework allows secure, inter-domain measurements, thanks to a new methodology purposely developed. To overcome the problem of measuring QoS data in Autonomous Systems not under the direct control of the initiator of the measurement, an XML file is sent over to AS being visited by the packet flow. The local device will then measure the QoS of its domain, and recursively send a modified XML file to the following AS, until the final destination is reached. All the developed components were integrated, and locally tested. As part of the expected result from 6QM project, a knowledge base and a set of guidelines has been created, that could be exploited by operators and ISPs to meet the client demand in IPv6 advanced services with guaranteed and differentiated QoS. In particular, we refer to deliverables 1.5, "IPv6 QoS Measurement Guidelines for ISPs", and 2.9, "Guidelines for IPv6 measurement in home environment". Extensive dissemination and Linkage with other related Fora and Projects has been carried out, and allowing an outstanding progression in the standardization activities, in order to solve some remaining R&D issues and disseminate the project results. The standardization activities can be considered a great success of this project, having in IETF 9 draft, of which 2 already in the final stage. All these operational results are supposed to be helpful in giving an answer to market expectations concerning the development of the New Generation Internet based on IPv6 technology with a strong constraint on QoS.

The QoS measurement technology developed with in the 6QM project will provide key components to support the fast transition towards IPv6 based networks in Europe and worldwide. Given the impact and relevance of the issues addressed, the 6QM consortium considers the task of dissemination as highly important. This deliverable describes achievements by the project 6QM in the dissemination of project results into target communities. It encompasses the 2nd half of the project covering the period from November 2003 to December 2004. In addition it contains supplements of on-going activities performed during the finalizing period to April 2005, in particular this relates to standardization work, which will be continued by individual partners beyond the lifetime of the project. In response to the different information needs of each of its target customer groups the project has planned for a comprehensive communication strategy to ensure that awareness for project activities and an uptake of project results is achieved. The companion document D5.4.1 reported on the first project phase of the first 14 project months. In this document we describe how this strategy has been continued in the 2nd project period. The report summarizes material produced and actions undertaken to distribute results to the various audiences. It gives an account on presentations and demonstrations, established liaisons within the 5th Framework Programme and 6th Framework Programme, material accessible via the website, contributions to international standardization and results from the 6QM organized workshop event on IP measurement. Appended to the document are the partners' exploitation plans of project results.

This series of deliverables are the monthly reports of the project, indicating the progress made. All partners specify the deliverables and topics they have been working on, expected dates of delivery and achievements. This deliverable covers from the beginning of the project until the Month 28.

This document addresses the needs for interoperability of measurement solutions across service provider boundaries. It takes into account the need to measure end-to-end quality of service across administrative areas and over composite networks, and as such examines the exchange of measurement data between asynchronous systems. It provides recommendations for interoperability for following areas: - Collecting the information; - Reporting of the information; - Consolidation of the information; - Configuration; - Test plane.

The objective of this document is to give a brief overview of the ongoing standardization work at the IETF in the IPPM (IP Performance Measurement) working group. The working group general information is outlined, including the goals and milestones, as well as the current Internet drafts and RFC documents. The IPPM framework is defined, beginning with general measurement concepts, and then explicit definitions for all the metrics that have currently been defined. This set of metrics includes: - Metrics for Measuring Connectivity. Metrics that determine whether pairs of hosts (IP addresses) can reach each other form the basis of a measurement suite for connectivity. - Metrics for Measuring One-way Delay. The measurement of one-way delay from one host to another can be useful for a variety of reasons, some of which include: Certain applications do not perform well if end-to-end delay between hosts is large relative to some threshold value. Erratic variation in delay makes it difficult to support many real-time applications The larger the value of the delay, the more difficult it is for transport-layer protocols to sustain high bandwidths. - Metrics for Measuring One-Way Packet Loss. Understanding one-way packet loss from a source to a destination host can be beneficial for many of the same reasons as listed above for one-way delay. - Metrics for Measuring Round Trip Delay. Round-trip delay of a Type-P packet from a source host to a destination host. - Metrics for Measuring IP Packet Delay Variation. The definition of the IP Packet Delay Variation (ipdv) can be given for packets inside a stream of packets. The ipdv of a pair of packets within a stream of packets is defined for a selected pair of packets in the stream going from measurement point MP1 to measurement point MP2. The ipdv is the difference between the one-way-delay of the selected packets. Certain works are still in progress, including the definition of a one-way active measurement protocol, and an IPPM REGISTRY (list of all object identifiers for known RFC metrics), and an IPPM REPORTING MIB. The IPPM Working Group is the most appropriate place to submit standardization contributions related to IPv6 QOS measurement. This document identifies 2 contributions related to the 6QM project.

This document describes liaison activities already started by the 6QM project and provides a roadmap for the future inter-project cooperation. It extends on the approaches outlined in D5.1 Dissemination and Use Plan, which identified other IST projects and activities for potential co-operations. The following report provides an account on the liaisons activities performed during the last quarter since the publication of D5.1, and makes recommendations for further cooperation developing technical details on the targeted common research.

Rezultaty

Udostępnij tę stronę Udostępnij tę stronę w mediach społecznościowych

Pobierz Pobierz zawartość strony