IPv6 Pain Experiment

Masataka Ohta mohta at necom830.hpcl.titech.ac.jp
Fri Oct 4 20:45:26 UTC 2019


Matt Harris wrote:

>> That is called "provider lock-in", which is the primary reason,
>> when IPng WG was formed, why automatic renumbering is necessary for
>> IPv6.

> If this is a concern, then get an allocation from your local RIR and 
> announce it yourself. Then no provider lock-in based on address space
> of any sort.

So, IPv6 did not failed, because manual renumbering is easy and,
even if it is hard, we don't need CIDER.

Very convincing argument.

> In general any sort of provider move is going to be disruptive if you
> don't have your own address space,

Automatic renumbering is the technology to make it not disruptive.

It is doable, if the tricky part of nameserver address changes is
properly taken care of.

But, people who think renumbering just a prefix change can not
do it, resulting in garbages like A6 RRs or IPv6 itself.

						Masataka Ohta




More information about the NANOG mailing list