Hi, On Tuesday 26 January 2010, Vlad Ion wrote: > Unfortunately it is not as easy as just turning it on like in the case of > IPv4 where you set your ip and that's it - you have access to the > internet. If you set v6 instead of v4 on a server you get access to > people in the v6 and 6to4 by having a 6to4 relay but then you need nat-pt > for v4 > connectivity. or I may be mistaking but that was the case with the > servers I set v6 on till now. grantedly I come from a SOHO (customer side) background, but I don't really see where the difficulty is. What I normally do with networks that I switch to v6 is this: 0a) Leave IPv4 the hell alone! It works. Don't fix it. Don't even think about taking it away from the admins! They would eat you for lunch (or sooner). 0b) Don't plan for production use immediately. Plan for a transition phase of at least 6 months. This gives you time to explore your options. 1) switch v6 on (configure routers, there are easy HOWTOs) 2) add v6 to DNS and some other services (over time) 3) relax, your main job is done, most hosts will switch it on magically. The actual work time for this is very little. And you'll have a working dual-stack environment afterwards. Now you can start to transition applications one by one over a longer period of time. There are a couple of other things you can think about while the test and transition phases are still running - like unportable applications, address stability or renumbering after a provider switch, DHCP-PD, a final v6-only network in 2015 or maybe 2020, etc. There is really no need to disrupt the network or the global addressing architecture just because someone wants to replicate the old v4 work instructions on a v6 network - there is time enough to explore and create new instructions while the network already works in test mode. Konrad
Attachment:
signature.asc
Description: This is a digitally signed message part.