Further IPv6 rollout
With the initial rollout of IPv6 now well underway in the Computing Department, the next logical goal for IPv6 deployment within the University will be to try to improve the IPv6 user-base by increasing the number of IPv6-enabled hosts within the network. There are two logical ways in which this could potentially be pursued in our case: either by enabling dual stack IPv6 on the student network (ResNet) or in the University computer laboratories. Either (or both) of these can be done relatively easily depending on the requirements of ISS for managed IPv6 deployment.
If it is determined that initial IPv6 deployments should be managed then the lab networks are the natural choice as they are heavily managed by ISS whereas the student networks are much less so and so IPv6 can be enabled but left to the students to use IPv6 if they wish. Since the majority of hosts on the both networks will be MS Windows-based, the majority of new IPv6 users here are likely to be Windows XP-based but in the student network case ISS should be prepared for both Linux and Unix support also. Either of these deployments would potentially introduce several hundred new IPv6- enabled hosts to the network (at a conservative estimate) and so would greatly increase the scale of IPv6 deployment within the University campus.
Protocol Selection Policy Definition
With a significant IPv6 deployment in place, an issue that gains relevance at this point is in defining how IPv6 is used within IPv6-enabled hosts. The conventional choice within IPv6 early-adopters is currently not to use IPv6 as the first-choice protocol except in explicit special cases where IPv4 is not supported. This should certainly still be the case in the majority of our deployments where the network is dual stack and IPv6 is not yet able to offer an equivalent service to the existing IPv4 network.
Ultimate Goals
Once this is complete, IPv6 will be available across the entire network and so the ‘deployment’ will be largely complete. As such, it will then be possible to consider what other services could potentially be deployed across the infrastructure and while it is too early to define these with a great degree of confidence, multicast and mobility are both strong candidates for this as they feature strongly as IPv6- only applications that would showcase the potential of the new protocol.
