Why Bother?
With all of the things that can go wrong with IPv6, you might be wondering if you should be even thinking about deploying it. This is a legitimate concern, of course, but the alternative to deploying it now is deploying it later. There is no option to avoid IPv6 deployment completely.
As I said at the start of the last article, IANA has already assigned all of the IPv4 address blocks that it has. These will then be assigned to large ISPs, and then further assigned to their customers, and some time quite soon someone will request an IP address assignment and find that the network that they connect to has no more to offer.
At this point, they will have to start using IPv6 only. This is likely to happen in Asia first, because they got fewer IPv4 addresses to start with (and a lot fewer per capita) so if you need to communicate with people on that side of the world, then IPv6 is likely to be more important than if, for example, all of your business is inside the USA.
At this point, there will be no safety net of dual stack; some endpoints will be IPv6 only, and if you are IPv4 only then you will not be able to communicate with them. If you start deploying IPv6 now, then you can work out the problems while most of the hosts that you communicate with still have IPv4 support to fall back on.