On 4/25/06, <b class="gmail_sendername">Howard, W. Lee</b> <<a href="mailto:Lee.Howard@stanleyassociates.com">Lee.Howard@stanleyassociates.com</a>> wrote:<div><span class="gmail_quote"></span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
I've made some changes to this message:<br>I removed other mailing lists. I'm not a subscriber.<br>I changed to plain text formatting.<br>I tried to focus on ARIN public policy proposals.<br><br>Matthew Petach wrote:<br>
> > dynamicity. Huston's study indicated that there are folks whose BGP<br>> > announcements flap (due to TE) intentionally 1000's of times a day.<br>> BGP flap dampening is already well understood for limiting the impact
<br>> of flapping routes on your CPU, if that's a concern; it has no bearing<br>> on address allocation policy decisionmaking.<br><br>Dampening works, for the value of "works" which equals "suppressing new
<br>(better?) information about the path to a network," which is often equal<br>to "misrouting data because a new path was not calculated." If,<br>following<br>this proposed policy, aggressive dampening becomes commonly required by
<br>operators to maintain their networks, it might have a bearing on support<br>for this policy.</blockquote><div><br>Concern was raised about end networks changing routing state too quickly<br>for modern CPUs to handle; the way we handle that today is to dampen
<br>successive transitions once they reach a certain threshold level. I see no<br>reason why the current mechanism in use today in the IPv4 world should<br>not work just as well on the IPv6 network.<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;">
> pool. The policy cannot allow itself to be shaped by the<br>least-capable<br>> routers, or we'll be chaining ourselves to the past,unable to make<br>adequate<br>> forward progress so long as there's any network out there that's still
<br>running<br>> old hardware that's unwilling to upgrade.<br>> I agree we need to be reasonable--but please, let's not "dumb down"<br>the<br>> v6 internet just because some people aren't willing to step up to the
<br>plate and<br>> upgrade their routers when needed.<br><br>For values of "unwilling" approaching "unable." As the number of<br>prefixes<br>approaches infinity, and the beta cycle (or dynamism) approaches zero,
<br>the<br>CPU required will approach infinity. Even if you decide to recalculate<br>routes once a month, memory requirements increase, potentially beyond<br>what<br>current or foreseeable routers can handle. Yes, this is a reductio ad
<br>infinitum argument; the trouble is that there is not common agreement on<br>what the break point is, or how fast it might approach.<br><br>If it takes six months of testing to be confident in a new core router,<br>and you replace 5% of your routers every week, and you ignore all other
<br>maintenance, you could replace your core in a year. Another year for<br>distribution and edge. Any other kinds of devices take another year,<br>for testing and rollout. That's assuming you employ as many engineers
<br>as needed, and ignore other work.<br><br>I certainly don't mean you have to agree that these issues should be<br>paramount in policy-making, but I think that before being ignored, it<br>should be acknowledged that otherwise competent operators potentially
<br>have a serious problem.</blockquote><div><br>The point I was making is that if we limit IP allocation policies<br>based on hardware upgrade cycles, we shackle ourselves to <br>the ISP that decides that they should be able to run just fine
<br>with their IGS or AGS+, and therefore no additional prefixes<br>can be allocated because they're unwilling _or_ unable to<br>upgrade.<br><br>If we start putting limits on what can be assigned due to <br>concerns about what some providers may be unwilling to
<br>upgrade in their core, are we also going to entertain <br>proposals to limit IPv4 allocations so that those selfsame <br>providers won't have to upgrade their IPv4 routers either?<br><br>Let's be realistic. We don't have any limits on how much
<br>IPv4 deaggregation we'll support, we just trust the routers<br>to scale up to support the growth--why are we attempting<br>to treat IPv6 differently? Do we not expect IPv6 to take <br>over for IPv4 in the forseeable future?
<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;">> Provider independence is here already, period. It's too late to try<br>to put
<br>> the horse back in the barn<br><br>I don't understand what you mean here. Do you mean "policy proposal<br>2005-1<br>is in last call and I believe it will be approved"? The last call part<br>of<br>the process exists for a reason; the AC will review comments about
<br>2005-1<br>before deciding whether to promote the proposal. Or are you talking<br>about other regions, which set their own policies?</blockquote><div><br>I mean 'In the IPv4 internet of today, multihoming is a fact of
<br>life, and is expected by businesses. If IPv6 is ever intended <br>to be more than an academic mastubatory exercise, the<br>policies associated with IPv6 allocations will need to support<br>a similar or better level of provider independence for businesses
<br>as the existing IPv4 allocation policies.'<br><br>To that end, yes, I support 2005-1. I anticipate it being approved. <br>If it is not approved, and none of the other possible proposals<br>for IPv6 multihoming for enterprise networks are approved, I
<br>see no reason to think that IPv6 will ever be used as anything<br>other than an experimental protocol utilized by small groups<br>of researchers.<br> </div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
> Trying to stop provider independent addressing in IPv6 is simply<br>another way<br>> of saying 'IPv6 addressing is broken, let's just stick with IPv4<br>instead' -- because<br>> that's what the practical outcome is.
<br><br>Are you asserting that IPv6 is, or is not, "broken"? Is a PI policy<br>like 2005-1 intended to "fix" IPv6 or is it simply meeting the original<br>requirements from the IETF?</blockquote><div><br>
IPv6 without provider independent allocations is broken,<br>and will never be widely adopted. 2005-1 is one possible<br>suggestion on how to fix it. I support 2005-1 as a well<br>thought out means to fix an otherwise broken protocol.
<br></div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">> Any addressing scheme that tries to<br>> deny the need for provider independent addressing is less capable for
<br>> businesses than what they have today in the IPv4 world, and will<br>therefore<br>> not be used.<br><br>Interesting assertion. In the context of 2005-1, which provides<br>specific<br>"bridles" to use your metaphor, are you suggesting it should move
<br>forward,<br>or not? Whether it is too permissive or too restrictive is another<br>interesting question, but doesn't sufficiently inform the process.</blockquote><div><br>Yes. I have previously stated on other threads, and will
<br>continue to state that 2005-1 should move forward, or if<br>there is sufficient concern about wording, that one of its<br>progeny or siblings should move forward; but without<br>some form of provider independent allocation scheme
<br>for non-ISP multihomed networks included in the IPv6<br>allocation policy, IPv4 will continue to be the dominant<br>network platform, and no level of critical mass will be<br>reached to support widespread IPv6 deployment.
<br></div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">> So. Given that PI allocations are going to happen in IPv6 just as<br>they have in
<br>> IPv4,<br><br>When you say, "just as they have in IPv4," what do you mean? Do you<br>mean,<br>"any applicant in the first ten years will get one with little<br>justification,<br>and anyone after that will have to provide documentation of their need"?
<br>2005-1 isn't identical to IPv4, but uses IPv4 as its basis.</blockquote><div><br>I think 2005-1 is sufficiently similar to the current IPv4 allocation<br>system for my language to be applicable.<br></div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
> I *do* agree that stipulating that only the largest possible aggregate<br>assigned<br>> to a given AS *should be* announced by the AS is a reasonable addition<br>to<br>> the policy to help encourage aggregation, and prevent stupid routing
<br>mistakes<br><br>In this context, do you oppose 2005-1 since it lacks such a<br>recommendation?<br>If this is an RFC2119 "should" then can the policy stand without it?</blockquote><div><br>No, I don't oppose 2005-1 as it stands. I'm tossing out an idea
<br>for an addition that might help those who live in fear of route<br>table expansion feel more comfortable with it.<br>Strictly speaking, though, even with no changes, no amendments,<br>no adjustments, I think 2005-1 is more than adequate, and can
<br>be approved as it stands without fear of the internet melting down<br>due to its adoption.<br></div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
> I don't see why the discussion is dragging out for so long. This<br>isn't rocket<br>> science. Let's just nail the policy down, and move on with more<br>productive<br>> work.<br><br>Is that a statement for or against 2005-1?
</blockquote><div><br>Yes. I am hoping to see IPv6 become something other than<br>a theoretical protocol developed as an amusing experiment;<br>therefore, I support 2005-1.<br></div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
As to your statement about rocket science. . . developping consensus<br>among<br>people with opposite perspectives is distinctly difficult. I've never<br>tried rocket science, but I know of some nuclear physicists who find
<br>consensus-building to be challenging.</blockquote><div><br>That's nice. If we don't come to consensus and move<br>forward, this rocket is never going to lift off. And if IPv6<br>doesn't take off, then we'll see some *real* challenges
<br>when IPv4 space starts getting scarce, and the black<br>market for v4 space starts to solidify.<br></div><br>You're worried about routing table growth in IPv6? Imagine<br>what it'll be like when v4 space runs out in 5-8 years and
<br>suddenly those nice class A aggregates are being sliced<br>and diced into /22 sized chunks for everyone screaming<br>for provider independent, routable space. Suddenly,<br>supporting PI space in IPv6 won't seem that bad.
<br>The same routing table slots are going to get used,<br>period; all we're discussing here is whether we set<br>up the policy such that they can be used for IPv6<br>routes, or if we force them to be used for IPv4 prefixes.
<br><br>I'm supporting 2005-1 because I'd prefer to see those<br>prefixes handled on the IPv6 internet. It sounds like<br>you'd prefer to say 'no' to IPv6 PI space, and instead<br>have your routing slots consumed by IPv4 prefixes
<br>instead, and that's certainly your choice. I'm just<br>trying to point out why I think it's a more short-sighted<br>choice that will end up locking us more tightly into the <br>IPv4 world, and why I favour the other choice, which is
<br>to support 2005-1 or one of its ilk, and help bring the<br>IPv6 internet into the current millenium.<br><br>Matt<br><br></div>