The aim of P1113 was to provide practical recommendations for network planners on selected aspects of IPv6. This was done by functional experimentation with available IPv6 products. The Tsunami project focused on five issues:

Multi-Protocol Label Switching (MPLS) core networks provide means of traffic engineering and Virtual Private Network (VPN) services provision. Over the last years, many ISPs and telecom operators have invested in setting up MPLS networks. With IPv6 the question emerged, what kind of update, upgrade, or transient solution will enable MPLS core networks to transport IPv6 traffic. There is a range of different techniques available, from simple tunnelling to vendor specific solutions. A guideline for the usage of IPv6 over MPLS has been described by P1113, and the pros and cons have been compared. The type of usage of IPv6 in MPLS backbones will have implications on costs, network management, and flexibility, depending on the existing network environment. The comparison prepared by P1113 can help to make the right decisions.

The extended address space is one of the main advantages of IPv6. Address space can be allocated by the Regional Internet Registries (RIPE in Europe). The discussion about ideal allocations of addresses for various types of networks is ongoing. In Tsunami, typical network scenarios — large ISP, mobile operator, large corporate network and dial-up networks — have been outlined, and principal addressing guidelines for these scenarios were described. Special addressing cases were considered and focal attention was given to automatic network renumbering — we performed practical tests with stateful and stateless renumbering techniques.

In the area of DNS (Domain Name System), we looked at operational issues during the transition. In the transition phase from IPv4 to IPv6, a mixed environment of IPv4 and IPv6 will exist for a long period of time. In this period, the DNS servers will evolve from an IPv4-only state into a dual-stack server, and potentially at the end into a pure IPv6 stack. In every state a server should be reachable by the specific transport protocol: IPv4, IPv6, or both. To achieve this, various mechanisms could be used. One of the experimental ideas is DNS bridging; DNS servers forward requests from IPv4 to IPv6 and vice versa. In Tsunami, we looked at DNS bridging, but we had to realise that currently available DNS software implementations (BIND9) do not support practical DNS bridging with IPv6, therefore we could not perform practical experiments.

There are two more features of IPv6 besides the larger address space that are often considered as the big advantages of IPv6: The enhanced mobility support and the inherent security support (IPsec). Together, they enable a better support for nomadic Internet users than IPv4. In the context of the project, a Europe-wide, Ipsec-based IPv6 VPN has been established between the project participants for practical experimentation with several IPv6 IPsec implementations. The available implementations have shown that the technology is ready for use. Secure mobility will certainly become a common feature of IPv6 applications, which is not available on the IPv4 Internet today.

The project task with the greatest impact so far has been the continued work in the Next Generation Transition working group of the IETF. In the framework of the Tsunami project, discussions on the mailing list with other active members of the IETF as well as a presentation at the Minneapolis IETF meeting and input for the Yokohama meeting have been provided.

One of the main concerns of the project participants was that a plethora of transition mechanisms has potentially severe implications on operational aspects of the transition process from IPv4 to IPv6, particularly for network operators. Additionally, there are technical concerns about the interaction of all mechanisms, which has not been investigated by the IETF in depth. Meanwhile, the IETF has recognised that continued work on – partially academic – transition mechanisms could slow down the transition phase from IPv4 to IPv6. Therefore, it was decided in mid-2002 that a new working group should push a stronger consideration of operational IPv6 aspects. The 'v6ops' working group was started in the IETF, taking over some work of the 'NGTrans' working group, but giving stronger emphasis to operational aspects. While the Tsunami work was certainly only a minor contribution for stronger consideration of operators’ concerns, this development in the IETF is certainly a progressive step from an operator’s point of view.

This article cannot outline all the achievements of the project, as further work was done e.g. on routing aspects and Internet telephony. The recent announcements and activities from software and hardware vendors, service providers, and the project participants indicate that the introduction of IPv6 will speed up in 2003. The experience gained and the practical guidelines elaborated by the Tsunami project team are available to help the project participants’ companies and all Eurescom shareholders to master the challenges of the arrival of the tidal wave of IPv6.

More information on P1113 'Tsunami' and all project reports are available partly in web format and partly as downloadable documents at: www.eurescom.de/public/projects/P1100-series/p1113