<html><head><meta http-equiv="Content-Type" content="text/html charset=us-ascii"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;"><br><div><div>On Jul 17, 2014, at 09:28 , Gary Buhrmaster <<a href="mailto:gary.buhrmaster@gmail.com">gary.buhrmaster@gmail.com</a>> wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite">On Thu, Jul 17, 2014 at 3:35 PM, Mike Burns <<a href="mailto:mike@iptrading.com">mike@iptrading.com</a>> wrote:<br>....<br><blockquote type="cite">Others have noted we are rearranging the deck chairs on the Titanic with<br>these IPv4 policy issues.<br>Maybe it is time to metaphorically sweep those chairs into the sea so we can<br>move on.<br></blockquote><br>Perhaps that is an interesting idea. Remove needs testing for<br>IPv6 /48s, and let everyone move on to being able to access<br>all of the Internet. I would have to consider the other implications,<br>but I think I could get behind removing needs testing for /48's<br>given how large the remaining IPv6 address space is, and it<br>would be an interesting experiment and insure that there is<br>absolutely no reason that one cannot move to IPv6 now.<br></blockquote></div><br><div>I don't support removing needs testing for IPv6 /48s at this time, but I will point out that the existing needs test for IPv6 boils down to:</div><div><br></div><div>1.<span class="Apple-tab-span" style="white-space:pre"> </span>I have a site.</div><div>2.<span class="Apple-tab-span" style="white-space:pre"> </span>I want to connect that site to the internet.</div><div>3.<span class="Apple-tab-span" style="white-space:pre"> </span>The address for that site is within the ARIN region.</div><div>4.<span class="Apple-tab-span" style="white-space:pre"> </span>That site meets one or more of the following criteria:</div><div><span class="Apple-tab-span" style="white-space: pre;"> </span>A)<span class="Apple-tab-span" style="white-space: pre;"> </span>Already has IPv4 from ARIN (6.5.8.1.a)</div><div><span class="Apple-tab-span" style="white-space: pre;"> </span>B)<span class="Apple-tab-span" style="white-space: pre;"> </span>Will be multihomed (6.5.8.1.b)</div><div><span class="Apple-tab-span" style="white-space: pre;"> </span>C)<span class="Apple-tab-span" style="white-space: pre;"> </span>Will have 2000 host addresses in use within 12 months (6.5.8.1.c)</div><div><span class="Apple-tab-span" style="white-space:pre"> </span>D)<span class="Apple-tab-span" style="white-space:pre"> </span>Will have 200 subnets within 12 months (6.5.8.1.d)</div><div><span class="Apple-tab-span" style="white-space: pre;"> </span>E)<span class="Apple-tab-span" style="white-space: pre;"> </span>Has some other technical justification for needing a direct assignment rather than getting LIR assigned space (6.5.8.1.e)</div><div><br></div><div>It's not quite a Pez dispenser, but it's about as close as I would be comfortable making it.</div><div><br></div><div>From NRPM section 6.5.8:</div><div><br></div><div><h6 style="margin: 0.5em 0px; padding: 0px; border: 0px; font-family: arial, helvetica, sans-serif; line-height: 16px; background-color: rgb(255, 255, 255); font-size: 11.5px !important; position: static; z-index: auto;"></h6><blockquote type="cite"><h6 style="margin: 0.5em 0px; padding: 0px; border: 0px; font-family: arial, helvetica, sans-serif; line-height: 16px; background-color: rgb(255, 255, 255); font-size: 11.5px !important; position: static; z-index: auto;">6.5.8.1. Initial Assignment Criteria</h6><div style="margin: 0.5em 0px; padding: 0px; border: 0px; font-size: 1.2em; line-height: 1.5em; font-family: arial, helvetica, sans-serif; background-color: rgb(255, 255, 255); position: static; z-index: auto;"></div><div style="margin: 0.5em 0px; padding: 0px; border: 0px; font-size: 1.2em; line-height: 1.5em; font-family: arial, helvetica, sans-serif; background-color: rgb(255, 255, 255); position: static; z-index: auto;">Organizations may justify an initial assignment for addressing devices directly attached to their own network infrastructure, with an intent for the addresses to begin operational use within 12 months, by meeting one of the following criteria:</div><ol class="alpha" type="a" style="margin: 0.5em 0px; padding: 0px 0px 0px 30px; border: 0px; font-size: 10px; list-style: lower-alpha; font-family: arial, helvetica, sans-serif; line-height: 16px; background-color: rgb(255, 255, 255); background-position: initial initial; background-repeat: initial initial;"><li style="margin: 0.25em 0px 0px; padding: 0px; border: 0px; font-size: 12px !important; line-height: 1.5em; background: transparent;">Having a previously justified IPv4 end-user assignment from ARIN or one of its predecessor registries, or;</li><li style="margin: 0.25em 0px 0px; padding: 0px; border: 0px; font-size: 12px !important; line-height: 1.5em; background: transparent;">Currently being IPv6 Multihomed or immediately becoming IPv6 Multihomed and using an assigned valid global AS number, or;</li><li style="margin: 0.25em 0px 0px; padding: 0px; border: 0px; font-size: 12px !important; line-height: 1.5em; background: transparent;">By having a network that makes active use of a minimum of 2000 IPv6 addresses within 12 months, or;</li><li style="margin: 0.25em 0px 0px; padding: 0px; border: 0px; font-size: 12px !important; line-height: 1.5em; background: transparent;">By having a network that makes active use of a minimum of 200 /64 subnets within 12 months, or;</li><li style="margin: 0.25em 0px 0px; padding: 0px; border: 0px; font-size: 12px !important; line-height: 1.5em; background: transparent;">By providing a reasonable technical justification indicating why IPv6 addresses from an ISP or other LIR are unsuitable.</li></ol></blockquote><div><br></div></div><div><br></div><div>This is the official wording, but I think I captured the essence above.</div><div><br></div><div>Further, 6.5.8.1 goes on to provide several examples that should qualify under 6.5.8.1(e), so an inability to justify an IPv6 /48 for your site is, at this point, pretty much a failure of imagination or an impressively small site with no directly assigned IPv4 resources.</div><div><br></div><div>Seriously, how hard is it to stand up 2000 interface addresses over the course of 12 months? If I _REALLY_ had to do this, I bet 15 minutes and a PERL script could do it on a single host. (No, I'm not advocating that, just pointing out that short of an incredibly silly trivial infrastructure, this is really a no-brainer).</div><div><br></div><div>Remember, if you use privacy addressing, you're probably looking at 4-5 addresses per host minimum, depending on how you set your preferred and valid lifetimes and your refresh rate on rotating privacy addresses. So 2000 addresses is more like 400 active hosts, including all your routers, virtual machines, switches, web servers, etc. You can also allocate an IPv6 address to every service on a box if you want. (completely legitimate and good technical reasons to do so in some cases).</div><div><br></div><div>2000 host addresses is a _REALLY_ low barrier to entry, as is 200 subnets.</div><div><br></div><div>Owen</div><div><br></div></body></html>