Re: [v6ops] New Version Notification for draft-yourtchenko-ra-dhcpv6-comparison-00.txt (fwd)

[Mark ZZZ Smith <[e-mail]>]

----- Original Message -----
> From: Fred Baker (fred) <[e-mail]>
> To: Owen DeLong <[e-mail]>; Andrew Yourtchenko (ayourtch) <[e-mail]>
> Cc: "[e-mail] WG" <[e-mail]>
> Sent: Thursday, 19 December 2013 6:18 AM
> Subject: Re: [v6ops] New Version Notification	for	draft-yourtchenko-ra-dhcpv6-comparison-00.txt (fwd)
> 
>>>>  I don't think RAs are the single reason some people are 
> resisting deploying IPv6 at all. I think some people are resisting deploying 
> IPv6 because it needs to do one or both of two things for them - solve a 
> foreseeable problem that will effect them, or provide a tangible benefit. If it 
> won't do either of those things, then they don't currently see any value 
> from the effort involved. What are foreseeable problems or tangible benefits is 
> completely up to the perspective of the decision maker. Eventually they'll 
> see value in it, and deploy it. In some cases, some people who could deploy IPv6 
> may never see any value in deploying it, so they never will.
>>> 
>>>  It's not binary. You can represent as a number both the amount of 
> pain that a given problem (or set of problems) causes, and the amount of pain 
> that the big change in the network like adding a new protocol causes.
>>> 
>>>  As soon as the first number is smaller than the second one, nothing 
> gets done.
>>> 
>>>  Judging by the feedback I get from talking with people, even the 
> possibility of eventually eliminating RAs from their network might make the 
> second number smaller for a nontrivial amount folks, whose situation dictates 
> the DHCPv6.
>> 
>>  I find this very hard to believe. RAs are virtually automatic and free in 
> 99% of cases. If you feel the need to tune them, then that's pretty trivial 
> if you know enough to tune them properly.
> 
> Speaking as a pretty random individual.
> 
> First, I don't think that the statement "they never will" holds. 
> They indeed never will while the equation remains the same - IPv6 costs them 
> something and doesn't give them a measurable ROI, or enough of one. The 
> premise of an adoption curve is that the equation indeed changes - the cost of 
> not deploying eventually exceeds the cost of deploying, and as a result people 
> deploy. The question here is not the current state, it's the eventual state. 
> The data tells me that the equation is slowly changing. If we don't believe 
> that, we're all wasting our time.
> 
> As to the RA vs DHCP dispute, I think it's largely religious. If you live 
> in, or choose to live in, an RA/SLAAC world, and the information elements in the 
> RA meet your needs, by all means use it. If DHCP does something an RA 
> doesn't, and your network needs it, you're going to use DHCP for at 
> least that purpose. To the extent they are functionally interchangeable  - you 
> can get an address from either, or advertise the address of the DNS server, or 
> etc -  its the cook's choice. 
> 
> The primary argument for the RA, from my perspective, is that it identifies a 
> preferred router on a LAN for a prefix that it is advertising, and in the 
> unusual case in which a large number of devices simultaneously join a subnet 
> (such as when the subnet is first turned up or a network is renumbered), or if 
> some other attribute needs to be assigned to every device on a LAN in multicast 
> fashion, it has improved scaling characteristics. The primary argument for DHCP 
> is control - you can say what you mean to an individual system and have that not 
> need to be the same as you say to another, and if you have a requirement to know 
> for sure what system is using what address (for example for Source Address 
> Verification), you own the table.
> 
> If it's your network, it will be about the policies your network follows. 
> There is no inherently "right" answer about policy, and as a result 
> there is no inherently "right" answer about the implementation of 
> policy.

I'd suggest the above statement is really "If it's your network and hosts, it will be about the policies your network and hosts follows." I'm sure hosts were implied, however this issue very much depends on host implementations and host feature deployment, so I think it is worth specifically calling the hosts out.  

The trouble is that the rapid rise of the BYOD/"cloud computing"/smartphone/tablet model means that hosts and their IPv6 stack implementation capabilities are rapidly becoming less controlled and less able to be dictated by the people operating the network, in networks where this level of control is currently even  possible. I think we're seeing the split between who operates the hosts and who operates the network significantly and rapidly widen.

So the network operator might want to only use DHCPv6 for everything, but if they need to facilitate connectivity for even just one host that comes along that doesn't support DHCPv6 for everything, they'll need to continue to provide and support RAs too. I think that means RAs will need to exist for at least 5 or perhaps 10 years for those networks where "DHCPv6 for everything" is the goal.

Regards,
Mark.













> _______________________________________________
> v6ops mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/v6ops
>