<div dir="ltr"><br><br><div class="gmail_quote">On Wed, Jun 25, 2008 at 12:04 PM, Chris Grundemann <<a href="mailto:cgrundemann@gmail.com">cgrundemann@gmail.com</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br><div class="Ih2E3d"><br>
</div>I think that there could be some compromise reached between forcing<br>
every node on someones network to be dual stacked and just giving out<br>
ipv6 space to everyone with no requirement for use.<br>
<br>
Maybe (as you suggest) ARIN should give out a /32 (or other block) to<br>
every AS, no questions asked. Then, based on the knowledge that they<br>
have the v6 space, you build some requirements for requesting more v4<br>
space. I think most will agree that requiring every host to be dual<br>
stacked would both provide the most benefit and also be virtually<br>
impossible to enforce. Maybe instead, the requirement should be<br>
simply advertising the v6 block. </blockquote><div><br>Are you suggesting give out /32 PI to every AS? <br><br>If so:<br> Please don't add 40-50k routes to the global internet routing table if people aren't actually using them. <br>
You might want to make a clarification that you verify that the ASN is still in use and the organization actually wants PI vs PA - before assigning it a /32. <br><br>If doing this through PA is also acceptable - you'd have to give larger allocations to ISP's. Their actually are some organizations that do not want to obtain and maintain their own IP space. <br>
<br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">This would require the organization<br>
to get some sort of ipv6 transit service and would thus encourage them<br>
to actually utilize it. </blockquote><div><br>No it would encourage them to route it - which is not the same as using it. <br> </div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
It would also help push the demand for v6<br>
transit. Another possible requirement is that the organizations<br>
public website have a AAAA record from the previously assigned /32 (or<br>
whatever size block). I think that these two requirements are easily<br>
measurable and would noticeably affect ipv6 adoption rates.<br>
<div class="Ih2E3d"></div></blockquote><div><br>This last idea, is not *so* bad ..in that it gives you a means to measure - and is along the lines of the 'demonstrate you are making some effort with v6' requirement that's been suggested. However it pushes enforcement to the part of the cycle when they come back for IP's. <br>
<br>Some considerations.. <br><br>I ask for IPv4 <br>I get IPv4 + IPv6 and get told to put our public website on IPv6 as well<br>I come back for more IPv4 - and then what? <br>Do I get denied because I have enough IP's with IPv6? <br>
What if my public website is available on v6 -- but I want more v4 for my customers?<br>Do I get denied because my public website isn't reachable by an IP in the IPv6 allocation I got? <br>What if it's reachable on some other IPv6 IP?<br>
What if I contract out my public website to some company who doesn't do IPv6 yet despite all my requests for them to do it... and the IP's I'm asking for are for my corporate network, or my customers? I have to change webhosting vendors in order to be able to have my website on v6, in order to be able to get more IP's for the rest of my business?<br>
<br> </div></div><br></div>