Mícheál Ó Foghlú's Weblog

The night the IETF turned off IPv4

This is a good summary of the experiment at a recent IETF to disable IPv4 for one hour to test the issues with using an IPv6-only network infrastructure.The night the IETF turned off IPv4

After working on the new Internet Protocol version 6 (IPv6) for a decade and a half, the Internet Engineering Task Force decided it was time to turn off the old protocol (IPv4 or just IP). So this is what they did for an hour on the network used at the IETF meeting in Philadelphia this week. Network traffic plummeted from some 30Mbps to around 3Mbps as the meeting attendees who had IPv6 enabled could now only get at IPv6-reachable destinations on the Internet. Leslie Daigle, chief Internet technology officer for the Internet Society, who coordinated the IPv4 outage, considers the outage a success.

Views on Network Management

This article links to a well argued draft paper Why Telcos Don’t Get Networks. Abstract:

If there is a sector of the economy that should embrace network-based thinking, it is telecommunications. Surprisingly, the opposite is the case. The leading firms building telecommunications and Internet infrastructure increasingly emphasize consolidation, hierarchy, and exclusive control, rather than collaboration and decentralization. Regulators are dismantling legal frameworks that once promoted openness and interconnection, in favor of misguided efforts to incent proprietary investment. And many scholars, even those challenging the current drift of policy and business models, embrace a static worldview that is a relic of earlier eras. Network-based strategies are thus hard to find today in the so-called "network industries," even as such ideas flourish in adjacent digital information markets. This chapter explores the origins of this paradox, describes its manifestations in the legal and business environment, and traces a more hopeful future.

Metcalfe’s Law: more misunderstood than wrong?

An interesting discussion of Metcalfe’s Law Metcalfe's Law: more misunderstood than wrong?

The industry is at it again–trying to figure out what to make of Metcalfe’s Law. This time it’s IEEE Spectrum with a controversially titled “Metcalfe’s Law is Wrong”. The main thrust of the argument is that the value of a network grows O(nlogn) as opposed to O(n2). Unfortunately, the authors’ O(nlogn) suggeston is no more accurate or insightful than the original proposal. ... The typical statement of the law is “the value of a network increases proportionately with the square of the number of its users.” ... The unit of measurement along the X-axis is “compatibly communicating devices”, not users. ... Title of graph: "The Systemic Value of Compatibly Communicating Devices Grows as the Square of Their Number"

The Network Neutrality Debate

This is a great rant that pokes holes in just about everyone's arguments: LXer: Network Neutrality and an Internet with Vision.

In recently-aired plans by telephone companies, content providers who are willing to pay extra would get their content delivered at a higher bandwidth. While it's easy to wax indignant over telephone companies' presumptuousness in deciding what packets should travel at what times, it's harder to step back and take in the economic issues driving the proposed change. And there are technical questions about it as well.

CAIDA IPv6 Topology Maps

I was interested to see that CAIDA had released an updated IPv6 topology map based on statistics gathered in January 2008. The graphs are a visualisation of ASs (Autonomous Systems) in the Internet based on the number of routes between them. ASs are usually controlled by a single owner, and thus this diagram tries to capture logical complexity of the network of networks that is the Internet.

These graphs create an interesting picture of the complexity of the Internet. For comparison have a look at the IPv4 topology map (multiple versions). If you look at these you'll see that the current IPv6 topology is as complex as the IPv4 Internet in 2000, at the height of the dot com boom. Of course the current IPv4 topology is more complex (latest shown currently is 2007).

Ever since I discovered these graphs I have thought that they were a great way to visualise the complexity of the Internet (IPv4 and IPv6).

More Undersea Cable Pictures

It’s 2008 — Do You Know Where Your Internet Cables Are? - GigaOM

Discussion of NAT (and IPv4 address deplation)

Nice to see that we are starting to think of the good things that NAt gave us, as well as planning the move to IPv6 where it will no longer be required. I think this post has it right, NAT served a purpose, now move on.

As IPv6 Deploys, Will We Look Back on NAT as the Ugly Step Sister or Unsung Hero?

Internet prediced to reach zettabyte (annual US traffic)

TDP - Estimating the Exaflood

From YouTube, IPTV, and high-definition images, to “cloud computing” and ubiquitous mobile cameras—to 3D games, virtual worlds, and photorealistic telepresence—the new wave is swelling into an exaflood of Internet and IP traffic. An exabyte is 10 to the 18th. We estimate that by 2015, U.S. IP traffic could reach an annual total of one zettabyte (1021 bytes), or one million million billion bytes.

Internet Undersea Cables (Guardian, UK)

SeaCableHi.jpg (JPEG Image, 1703x1037 pixels) - Scaled (92%)

Thanks to James Mernin.

IPv6 Address Added for Root Servers in the Root Zone

ICANN | IPv6 Address Added for Root Servers in the Root Zone

MARINA DEL REY, Calif.: The Internet Corporation for Assigned Names and Numbers today took another step along the path of deployment for the next-generation IPv6 Internet addressing system.

IPv6 addresses were added for six of the world’s 13 root server networks (A, F, H, J, K, M) to the appropriate files and databases. This move allows for the possibility of fuller IPv6 usage of the Domain Name System (DNS). Prior to today, those using IPv6 had needed to retain the older IPv4 addressing system in order to be able to use domain names.

"The ISP community welcomes this development as part of the continuing evolution of the public Internet,” said Tony Holmes, chair of ICANN’s Internet Service and Connectivity Provider Constituency. “IPv6 will be an essential part our future and support in the root servers is essential to the growth, stability, and reliability of the public Internet.”

Name server software relies on the root servers as a key part in translating domains like “icann.org” into the routing identifiers used by computers to connect to one another. In 2007 the ICANN Security and Stability Advisory Committee concluded that ICANN should move forward with the enhancement of the DNS root service by adding IPv6 addresses for the root servers.

As more and more devices connect to the Internet they require unique Internet Protocol (IP) addresses. The remaining free pool of unassigned IPv4 addresses is being depleted by the growth of the Internet. IPv6 is the addressing protocol that increases the unique IP addresses from the 4 billion available in IPv4, to more than 340 trillion trillion trillion.

“Today’s addition of IPv6 addresses for the root servers enhances the end-to-end connectivity for IPv6 networks, and furthers the growth of the global interoperable Internet,” added David Conrad, ICANN’s Vice President of Research and IANA Strategy. “This is a major step forward for IPv6-only connectivity and the global migration to IPv6.”

Further technical information on the move is available at http://www.iana.org/reports/root-aaaa-announcement.html

Nework Neutrality arguments rumble on in the US

In this article David Isenberg critiques Scott Cleland's assertion that telcos need to disciminate against certain traffic for reasons of network management The Network Management Excuse.

Ireland has 2nd higest percentage of IPv6 DNS entries

In this interesting analysis of actual IPv6 deployment (for http - web servers, smtp - email server and dns - domain name servers) Examining Actual State of IPv6 Deployment it seems that Ireland is doing quite well in terms of DNS with 14% of our sites having an IPv6 address - not a high score but 2nd only to Lithuania. We have don't fare so well in have 4.1% of our mail servers IPv6 enabled - above the median in 35th place, and 38% of our web servers enabled - below the median around 80th.

IETF NSIS

I am here in Amsterdam speaking at a Marcus Evans conference on 9th Annual Signalling Systems for Future Telecoms (Exploring the Future of Networks and Interconnect for an All-IP World), quite a mouthful but a really good event.

The first talk today was from Martin Stiemerling (NEC) who co-chairs the IETF nsis activity: Next Steps in Signaling (nsis) Charter. This is a fascinating development allowing flexible in-band signalling and QoS (Quality of Service) negotiation over IP routed networks.

As usual the main problem with any inter-domain QoS across the Internet is a business issue, rather than a technical one of standards: why would any network operator give some other operator's customers priority on their networks, so most QoS systems are applied within a single network autonomous system only, administered by a single operator. So just because I think my VoIP is important, doesn't mean every operator in my path will agree, even if the standards allow them to.

Jim Bound on IPv6

An excellent interview with Jim Bound, CTO of the IPv6 Forum: IPv6 guru predicts last-minute switch to protocol - Network World

Summary of IPv6 Announcements

An excellent summary of major announcments about IPv6 this year: The Year IPv6 Made it to Major League

May 6th 2007: ARIN board of trustees passes a resolution advising the Internet community that migration to a new version of the internet protocol, IPv6, will be necessary to allow continued growth of the internet.

June 29th 2007, Puerto Rico: ICANN Board resolution states that: The Board further resolves to work with the Regional Internet Registries and other stakeholders to promote education and outreach, with the goal of supporting the future growth of the Internet by encouraging the timely deployment of IPv6.

Oct 26th 2007 at the RIPE 55 meeting in Amsterdam: “Growth and innovation on the Internet depends on the continued availability of IP address space. The remaining pool of unallocated IPv4 address space is likely to be fully allocated within two to four years...the deployment of IPv6 is necessary for the development of future IP networks”

Nov 15th 2007: IGF meeting, Rio de Janeiro, chairman’s report: “Panelists also discussed the eventual exhaustion of ICANN’s reserve of unassigned IPv4 addresses.”

It was made clear that this would not cause the Internet to fail, but this was used to indicate the importance of the effort to bring the IPv6 network on-line and the need for the full interoperability between the IPv4 and IPv6 networks.

This is but a small sample of the fast growing visibility IPv6 acquired this year, 2007.

The internet ecosystem, ranging from network Operator Groups (NOG) to peering forums to Regional Internet Registries (RIR) to ICANN, to IANA and even the United Nations sponsored Internet Governance Forum (IGF), had IPv6 quite high on its agenda this year. This rather sudden surge has taken at least some of us, amongst long term IPv6 proponents and advocates, by surprise, as it took close to a decade, lots of persistence, lots of convincing and periods of doubt, to get to this point. This is a period those involved in the early IPv6 experiments and deployment in the Research and Education Networks, the IPv6 Forum since its beginnings back in 1999, should savor and enjoy. Not to mention those who fathered RFC’s, spent time at IETF and late nights on numerous mailing lists and never ending e-mail strings.

Cisco IPJ - lots of good IPv6 discussions

The most recent issue of the Cisco IPJ (Internet Protocol Journal), that is available for free on-line, has a number of very interesting articles on IPv6.

First Geoff Huston, APNIC updates us with his view of the big picture on IPv4 depletion, and the slow pace of IPv6 roll out. He is right to point out that a continued dual-stack strategy, that many seem to favour, will continue to require IPv4 addresses, and will thus become untenable when the IPv4 addresses themselves run out. There isn't much time to deploy IPv6 everywhere, within the IPv4 depletion timescale.

Then Iljitsch van Beijnum gives his views on IPv4 address consumption, trying to focus not just on the /8 block allocations from IANA to the RIRs, but on the subsequent allocation by the RIRs to ISPs and others.

Then Leo Vegoda talks about the use of unregistered IPv4 address space: "Many organizations have chosen to use unregistered IPv4 addresses in their internal networks and, in some cases, network equipment or software providers have chosen to use unregistered IPv4 addresses in their products or services." He discusses the potential problems that could ensue.

On a separate, but related theme, that of security for IP networks, Kunjal Trivedi, Cisco Systems and Damien Holloway, Juniper Networks discuss secure multivendor networks.

This issue of the IPJ could thus be seen as the definitive one in terms of summarising the key issues of IPv4 depletion, and IPv6 deployment. We have 2 years before the /8 IANA pool runs out, and a further 12-18 months before the RIRs then cannot comply with requests for new IPv4 addresses. I suggest everyone who is interested in the future of the Internet as it is now (not some abstract new network we haven't thought of yet) read these papers and become conversant with the arguments, and then start deploying IPv6 as quickly as possible.

IPv6 SHIM6 Interoperability Testing with Moonv6 in University of New Hampshire

Over the summer Irish National IPv6 Centre have conducted interoperability tests with Moonv6 in the InterOperabilty Lab (IOL) in the University of New Hampshire: UNH-IOL Tests IPv6 for the Enterprise.

The university lab partnered with the Waterford Institute of Technology in Ireland to extend the testing into less familiar territory. This portion of the testing focused on an innovation not possible with today’s Internet called Site Multihoming by IPv6 Intermediation, or SHIM6 for short. Of special interest in financial transactions, SHIM6 is an IPv6-only failover function that kicks in if one side of a link goes down, automatically rerouting the connection without affecting the download in progress.

IPv6 is the successor to the current IP infrastructure that underlies data in today’s Internet and enterprise networks. The new protocol greatly enlarges the pool of IP addresses needed to network new servers, laptops, phones, printers, etc. While some geographies have already run out of IP addresses, it has been predicted that North America will face IPv4 address space exhaustion between the years 2010-2012. IPv6’s increased address space is expected to make better use of emerging technology areas like VoIP, video and various interactive multimedia applications as well.

Other benefits touted for IPv6 include simplified network architecture, an increase in new services, and increased number of network nodes, built-in security, and the ability to "plug and play" devices that are IPv6 enabled. After first getting involved with the protocol in the late 1990s, the UNH-IOL has been actively testing and debugging IPv6 devices on the Moonv6 network since 2003.

This work is related to the EU FP6 IST project ENABLE that looks at many aspects of IPv6 mobility, and promotes suitable changes to the IETF standards.

Zimbie launches beta Instant Messaging client/server for automated IM-"bots"

Another TSSG spin-out company, Zimbie (not fully launched yet), has just released a public beta of their software that allows users to easily setup automated Instant Messaging "bots" (robots to automate tasks). The idea is that what appears to be another friend on your IM buddy list actually talks to a server that performs some task and returns the result as text in an IM chat. The Zimbie client itself can allow the bot to return more complex data than just text, but you don't have to use teh Zimbe client to access a bot, any IM reader will do. The aim of the beta release is to demonstrate the ease of setting up a bot without doing any programming.

The beta release has attracted some attion on the blogosphere: Blognation Conor O'Neill), Cloudlands (John Breslin), KillerStartups, and more.

Nubiq launches mobiseer

One of the TSSG spin-out companies, Nubiq, has recently launched a new product, mobiseer (mobiseer.mobi from your phone), to help mobile users bookmark and tag useful sites. The main advantage is that it is specifically targetted at mobile browsers. This product is in addition to Zinadoo that helps build mobile websites easily.

Here is a good review of the mobiseer beta release from Conor O'Neilmobiseer is ma.gnolia for mobiles.

Mapping the Internet

This is an interesting new approach to mapping the Internet Technology Review: Mapping the Internet. The work draws on the research project DIMES, and has some interesting interactive graphing tools, and a downloadable client to join in the mapping project.

It is interesting to compare this to the excellent work done at CAIDA, such as the AS-level topology maps of IPv6 and IPv4, using these tools (skitter probes, BGP table analysis, and ancillary tools).

Returning to the magazine article from Technology Review this extract gives a good flavour of the contents:

It's the first study to look at how the Internet is organized in terms of function, as well as how it's connected, says Shai Carmi, a physicist who took part in the research at the Bar Ilan University, in Israel. "This gives the most complete picture of the Internet available today," he says.

While efforts have been made previously to plot the topological structure in terms of the connections between Internet nodes--computer networks or Internet Service Providers that act as relay stations for carrying information about the Net--none have taken into account the role that these connections play. "Some nodes may not be as important as other nodes," says Carmi.

The researchers' results depict the Internet as consisting of a dense core of 80 or so critical nodes surrounded by an outer shell of 5,000 sparsely connected, isolated nodes that are very much dependent upon this core. Separating the core from the outer shell are approximately 15,000 peer-connected and self-sufficient nodes.

Take away the core, and an interesting thing happens: about 30 percent of the nodes from the outer shell become completely cut off. But the remaining 70 percent can continue communicating because the middle region has enough peer-connected nodes to bypass the core.

With the core connected, any node is able to communicate with any other node within about four links. "If the core is removed, it takes about seven or eight links," says Carmi. It's a slower trip, but the data still gets there. Carmi believes we should take advantage of these alternate pathways to try to stop the core of the Internet from clogging up. "It can improve the efficiency of the Internet because the core would be less congested," he says.

To build their map of the Internet, published in the latest issue of the Proceedings of the National Academy of Sciences, the researchers enlisted the assistance of 5,000 online volunteers who downloaded a program to help identify the connections between the 20,000 known nodes.

The closest I can find to a full academic paper on this, rather than the magazine style article linked above, seems to be based on slightly earlier work very similar to CAIDA's, is an AS level topology analysis (paper).

@MISC{carmi-2006,
  author = {Shai Carmi and Shlomo Havlin and Scott Kirkpatrick and Yuval Shavitt and Eran Shir},
  title = {MEDUSA - New Model of Internet Topology Using k-shell Decomposition},
  url = {http://www.citebase.org/abstract?id=oai:arXiv.org:cond-mat/0601240},
  year = {2006}
}

Abstract

The k-shell decomposition of a random graph provides a different and more insightful separation of the roles of the different nodes in such a graph than does the usual analysis in terms of node degrees. We develop this approach in order to analyze the Internet's structure at a coarse level, that of the "Autonomous Systems" or ASes, the subnetworks out of which the Internet is assembled. We employ new data from DIMES (see this http URL), a distributed agent-based mapping effort which at present has attracted over 3800 volunteers running more than 7300 DIMES clients in over 85 countries. We combine this data with the AS graph information available from the RouteViews project at Univ. Oregon, and have obtained an Internet map with far more detail than any previous effort. The data suggests a new picture of the AS-graph structure, which distinguishes a relatively large, redundantly connected core of nearly 100 ASes and two components that flow data in and out from this core. One component is fractally interconnected through peer links; the second makes direct connections to the core only. The model which results has superficial similarities with and important differences from the "Jellyfish" structure proposed by Tauro et al., so we call it a "Medusa." We plan to use this picture as a framework for measuring and extrapolating changes in the Internet's physical structure. Our k-shell analysis may also be relevant for estimating the function of nodes in the "scale-free" graphs extracted from other naturally-occurring processes.

I am going to try and find out some more about this very interesting work, and I'll update this posting to reflect new information as I receive it.

IMS (3GPP and ETSI TISPAN)

Telecoms.com – News, Events and research for the telecoms industry

It is good to see emerging consensus in the telecommunications standardisation community on the use of IMS for Fixed Mobile Convergence (FMC). Here it is reported that ETSI TISPAN and 3GPP have agreed how to proceed with the IMS Release 8 process.

ETSI TISPAN, the European standards body for the fixed-line half of the Next Generation Network world, this week agreed terms with mobile standards body, the 3GPP, on how to prevent the fixed and mobile versions of IMS from wandering apart.

IMS originated with TISPAN before being taken up by 3GPP for UMTS R99/5, but since then, the fixed-line and co-ax people have become more interested in the idea again. Both standards committees have been co-existing peacefully but perhaps not as happily as one might like.

The development of a Common IMS - common to fixed, cable and mobile networks - is going to go ahead in a 3GPP group, Services Specification SA-1, after a decision at last week's 3GPP plenary meeting in Busan, South Korea.

The job has to be done by the end of 2007 in order to be ready for the launch of UMTS Release 8, or "the IMS you can deploy" as some people call it.

Stephen Hayes, of Ericsson and 3GPP TSG-SA chair, said: "Over the next few months we must stabilise the Release 8 requirements and absorb the incoming Common IMS work."

IPv6 Alloactions to Enterprises in Europe - a problem?

In this excellent article Concerns grow over IPv6 migration - ZDNet UK the issues that effect the widespread adoption of IPv6 addressing within the enterprise networking community are addressed. It looks like the time has come to open up allocations more for enterprises and not just ISPs. I've worked with Tim Chown of the University of Southampton and the UK IPv6 Task Force on a number of activities, including the EU IPv6 Cluster -- a forum for interaction between various EU projects involved in IPv6. He has his finger on the pulse of the technical issues and has a good understanding of the business issues, so I respect his opinion on this matter. Note also the underlying message, as articulated here on a number of occasions, that even the conservative estimates of IPv4 exhaustion are converging on 2010 and 2011, not very far away.

Enterprise Ireland IMS Workshop

On the 5th June Enterprise Ireland hosted a workshop on IMS (the IP Multimedia Subsystem that is the standard at the heart of emerging data and voice in Fixed-Mobile Convergence for the telecommunications industry). The main guest speaker was our colleague Professor Thomas Magedanz of Fraunhofer Institut für Offene Kommunikationssysteme FOKUS, Berlin; he's also a visiting professor in Waterford Institute of Technology working with the TSSG on a number of Irish and EU funded research projects. Amongst other things in his presentation he was promoting his Open IMS Playground, which we use as an integrated testbed between the TSSG and FOKUS.

The TSSG has started a website to track the Enterprise Ireland ILRP (Industry-Led Research Programme) IMS ARCS, in which FOKUS are a partner, and a recent posting on that site summarises the purpose of the workshop, and gives links to all the presentations made EI IMS Workshop | IMS ARCS Project.

In fact I was originally due to share the platform and speak at this Enterprise Ireland workshop covering some background on the TSSG, and giving some initial details of the new IMS ARCS programme. Robert Mullins of the TSSG deputised for me and gave this presentation.

So watch this space for future news on the TSSG's research in IMS, or track the IMS ARCS project directly on its own website. RSS feeds of the blogs are available as well.

IMS ARCS is an Industry Lead Research Programme (ILRP) lead by the Telecommunications Software & Systems Group (TSSG) in Waterford. The IMS ARCS project is building a platform to facilitate the creation of IMS based services so that companies can tap into this market more quickly and easily, while ensuring consumers are not swamped by the resultant proliferation of services. As part of this work, a world class IMS Test Bed will be set up by TSSG.

The vision of the IMS ARCS project is an environment where end-users can be immersed in a world of diverse IP-based media, voice or data services that can be received off any number of network types. The IMS ARCS platform will allow services to be selected and personalised based on the everyday needs of the end-user and on their current context. Every user’s experience of the network and its services will be differentiated by the user’s defined requirements, their past usage and their current context.

These services will be supported by the provisioning of both existing and next generation network technology including both scalable and seamless integration of multiple heterogeneous networks, and ad-hoc and sensor networks. This will encourage and empower network and service operators to create new business opportunities and profitable services in an integrated mobile world.

IMS ARCS (EI ILRP) Partners:

• Telecommunications Software & Systems Group (TSSG), Waterford Institute of Technology
• Centre for Converged Communications Services (3CS), Waterford Institute of Technology
• Irish National IPv6 Centre, led by TSSG in Waterford Institute of Technology
• Hamilton Institute, NUI Maynooth
• Interaction Design Centre, University of Limerick
• Applied Software Research Centre, Athlone Institute of Technology
• Next Generation Network Infrastructures Competence Centre, Fraunhofer FOKUS

OMTP (Open Mobile Terminal Platform) Announces Requirements for IMS

The OMTP (Open Mobile Terminal Platform) announced on 12th June 2007 that it had completed its set of requirements that lay the foundations for the seamless deployment of IMS (IP Multimedia Subsystem) applications and services and pave the way for a more consistent end-user experience.

While many in the mobile industry believe that IMS will play a significant role in the future of data services, there are issues over the practicalities of deployment, which need addressing to enable IMS to truly deliver on its potential:

• The applications that will built on IMS must be able to consistently access any necessary services;
• The way in which applications respond upon receiving incoming IMS events must be defined;
• The end-user must have a consistent and coherent experience;
• Specifications must be consolidated and tiered for widespread uptake.

In its current form the IMS proposition fails to sufficiently address these four key areas and falls short of the full end-to-end experience that is required. OMTP’s requirements detail the steps needed to address these fundamental issues.

Tim Raby, OMTP CEO commented, “Despite the issues around its deployment, IMS delivers a host of potential benefits for operators, not least the ability to improve Quality of Service for end-users. For this to become a reality however there is a certain amount of ‘knitting’ needed to join up various elements. Our requirements give clear guidance to ensure IMS has the opportunity to deliver on its promises.”