Mindshift – Success Factors for Implementation of IPv6
10. January 2018
A sensible planning requires a long-term perspective. In many cases, a mindshift is necessary.
One of the important things to do when starting an IPv6 project and as part of a first overall assessment and overview of the current environment is to create an overview of upcoming IT initiatives for the next 3 – 5 years. Usually IPv6 is listed as an individual project, which has already been delayed several times, due to lack of time and budget or both.
Fact is, that IPv6 is a fundamental aspect of all other IT initiatives, simply because it is the transport protocol for all services. Just like security, it cannot be excluded and postponed to next year.
This approach has the advantage that the cost and budget for deployment of IPv6 can be spread over multiple years and budget items, as they are a basic part of each project.
Goal of an overall planning:
- Minimize cost and risk
- Carefully plan overarching concepts (architecture, addressplan, security, processes, vendor strategy)
- Educate teams, establish support channels, build knowhow and expertise internally and with partners
For the first version of the overarching planning and roadmap the following methodoly has shown to be good practice and successful.
Discover and strategy are the two areas that need to be carried out in advance and with the whole organization in mind. It is important to work on these items with crossdisciplinary inter-departmental teams. This ensures that all important aspects and affected organizational parts and IT initiatives are taken into consideration. With the alignment of the IPv6 implementation roadmap with other IT initiatives, a lot of cost can be saved. There are many organizations that state, that thanks to foreward-looking planning they had no additional cost for hard- or software for IPv6. They were able to use all the regular lifecycles of their products.
Architecture and roadmap
The following very much simplified schematic view shows an way to visualize high-level planning. Each of the four areas (client, backbone, server, Internet) can either be IPv4-only, dual-stack or IPv6-only. The challenge is in finding the leanest and least costly (also in terms of labour) path for the migration. Because due to cost reduction we try to sync this with other IT initiatives, this process can easily take 3 to 5 years. But this also has a great advantage. The migration becomes more digestable for the organization. It creates time to build expertise, fix bugs with vendors and integrate the learnings.
There is no best prescribed order. The order is the result of the individual environment, the IT strategy and the lifecycles. The overarching goal is to reach an IPv6-only status in as many parts of the network, for as many services as possible and as soon as possible.
Trends and best practices
For many years in the early days of IPv6 there was a consensus that dual-stack was the simplest and most elegant way of introducing IPv6, because it allows easy access to IPv4-only legacy applications as long as necessary. Organizations that adopted this path have often changed strategy after a couple months of operation. The complexity of operating and troubleshooting IPv4 and IPv6 in parallel was way too high. So today many strive to migrate to IPv6-only as soon as possible. Cisco built an IPv6-only campus in San José. They presented their experiences and learnings at an Swiss IPv6 Council event in october 2017. The engineer presenting who had built that campus summarized the experience as follows: “In 2017 dual-stack is a waste of time. IPv6-only works.”