Re: [v6ops] The bottom is /112 (was: RE: Extending a /64) -- How about new fixed bottom /80 win-win for all - epiphany at 6:54am after v6ops preso

[Gyan Mishra <[e-mail]>]

On Sat, Nov 21, 2020 at 12:48 AM Mark Smith <[e-mail]> wrote:

>
>
> On Sat, 21 Nov 2020, 09:27 Gyan Mishra, <[e-mail]> wrote:
>
>>
>> (top posting)
>>
>
>> As I would like to clear the air as well as get to the crux of the v6ops
>> presentation development results as well as next steps for this draft and
>> the 6man variable slaac solutions draft:
>>
>>
> You're asserting that everybody has accepted there is a problem to be
> solved, and that everybody has accepted that the only way to solve it is to
> change the default 64 bit IID size. This is both premature and not a
> representation about "everybody's" opinion you can make.
>

    Gyan> I believe after the v6ops presentation I gave on the problem
statement as well as the countless threads and interest I think more folks
think their is an issue to be solved then don’t.  I think the problem is
clear. I never said that this is the only way to solve the problem.  This
is just one idea of a possible solution and there definitely maybe others
that are better and I am not denying that.  Nothing premature as this is
just and idea of one possible proposed solution.

>
> I haven't accepted either of those things.
>
> I'm not happy with the idea of changing away for 64 bit IIDs because I
> don't think 64 bit IIDs are the problem. Plenty of operators have given out
> /56s. I'm working at my second one. It is not impossible.
>

    Gyan> All of the countess threads have all been spawned from the
variable slaac problem statement draft and solution to change the 64 bit
IID boundary.  There are two sides to the camp of changing versus not
changing as this idea to change is not new as it comes up every few years
with a fierce onslaught of threads for and against.  Their is definitely
interest in changing but thus far the balance of not changing as been the
weighted majority over not changing.  I am hoping this time around maybe we
have a game changer of a problem to be solved and the fixed 64 bit IID
change gets close to workgroup consensus.

>
> Many things below are incorrect, uninformed, or quite frankly ridiculous.
> Some read like you've let your imagination go wild and then have just
> assumed because you can imagine it it will then inevitably happen in the
> near future.
>

   Gyan> That is your POV.

>
>
> For example, DHCPv6 does not provide a prefix length to hosts. Your
> proposal does not make SLAAC "consistent' with what DHCPv6 does, because
> DHCPv6 doesn't work the way you think it does (look up IA_NA and IA_TA in
> the DHCPv6 RFC).
>

    Gyan> Understood.  I think you may have misunderstood what I meant and
maybe I did not state clearly and that’s my fault if the latter.  My point
is that DHCPv6 is length independent as the sever sends 128 bit managed
address M=1 to host. Point here is that both static and dhcpv6 are prefix
length independent and slaac is fixed at 64 bit.  This does introduce
incompatibility issue with slaac host endpoints not having parity with
dhcpv6 and static being prefix length agnostic. So as stated in the v6ops
deck this prevents operators from having longer then 64 bit prefix lengths
on host subnets due to the 64 bit boundary.  I understand that there are
many addressing guidelines that state that due to this parity issue that
/64 is recommended for all host subnets. However operators would like this
to change so that >64 prefix can be deployed to host subnets.

>
> "the idea of a wearable /48 will really be many /48s"? Do you really think
> we're going to have more devices on our person that we'll need more than 65
> 536 /64 subnets (a single /48) to number them? That sounds like science
> fiction to me for at least the next 200 years or more.
>

Gyan> I did not start the original thread on “wearable” /64.   Please see
6MAN archives.  I just took that and ran with it based on future network
infrastructure developments upcoming and paradigm shift and a reality of a
digital transformation with 5G in the context of “network slicing” which
will require exorbitant amount of address space.  It’s no different then
concept of cloud computing and hypervisor virtualization which required
many more addresses per VM as compare to physical servers.  So now apply
that to the wireless operators RAN front haul and backhaul xHaul
virtualization framework with “network slicing” game changer paradigm shift.

>
> IPv6 isn't and isn't required to be the final and only Internet protocol
> for the remainder of humanity, so we don't have to try to accommodate every
> conceivable idea that somebody can come up with that might cause the 128
> bit address space to be too small.
>
>
>>
>> This thread was in light of Lorenzo kindly pointing out that upgrading
>> 3GPP is not all that needs be done for Cameron’s 64share v2 to work - as
>> all mobile devices would stop working- as slaac would not would be
>> effectively broken.
>>
>> The mobile device would receive a shorter prefix let’s say /56 but not
>> know what to do with it since it’s expecting a /64.
>>
>> So that a major gap and the only solution is updating RFC 4291 removing
>> the 64 bit boundary allowing for shorter prefix and now as well longer
>> prefixes to work and in that respect now provide the much needed parity
>> with static and DHCPv6 which can do any prefix length.
>>
>> So that is a drastic change to RFC 4291.  However, in light of this
>> development on the v6ops 109 call, my balancing act of best of both worlds
>> and also to keep everyone happy to make this a WG effort for this change by
>> proposing in the subject heading /80 fixed boundary and not a variable
>> slaac change allowing all bits vlsm.
>>
>> Basically stealing 16 bits for network prefix out of the IID, still
>> keeping the fixed boundary so longer than 80 would NOT  be allowed.
>>
>> A /64 would now be equivalent to a /48 with now 64k /80’s.
>>
>> This /80 would keep the operators and law enforcement happy as now 16
>> bits less helps traceability but is still long enough for 48 bits of
>> privacy to IP correlation by attackers.
>>
>> This /80 would be a nice optimal balance as it would keep wired broadband
>> and mobile handset customers happy respecting their privacy as the 16 bits
>> less of heuristics is minimal change that will impact IP correlation by
>> attackers.
>>
>> The IID as it’s less than the current 64 bit cannot use MAC based EUI64
>> IID, which is not a problem as Mac based IID is not recommended as most all
>> manufacturers use RFC 4941 and I believe Linux flavors some use stable IID
>> RFC 7217.
>>
>> So now the 48 bit IID would require a random IID generation schema so can
>> use either RFC 4941 privacy extension or RFC 7217 stable IID to generate
>> the 48 bit IID.
>>
>> 3GPP subtending would now work issue mentioned in the problem statement
>> draft without even having to use 64share as now longer prefixes up to /80
>> would be supported allowing for further segmentation of downstream devices.
>>
>> This also would help wired broadband and soon fixed 5G broadband
>> proliferation where operators in light of BCP RIPE-690, are sill allocation
>> via BNG gateways a /64, now operators  can stay as-is, as the /64 would now
>> be allowed to be further segmented supporting 64k /80s, way more then
>> enough for SOHO.
>>
>> This would allow 64share if used by 3GPP operators to work and would not
>> require the 3GPP specification to be updated.  We don’t know even if the
>> 3GPP architecture specification can be updated to support shorter prefixes
>> and if the 3GPP consortium of operators would agree to it.  So that is all
>> theoretical of that change is possible.
>>
>> As with 5G with Enhanced VPN framework SR steering of high priority
>> traffic, traffic isolation and Network slicing capabilities becomes
>> mainstream and will soon be a real world reality and as fixed 5G broadband
>> proliferation takes off and mobile 5G == the idea of a wearable /48 will
>> really be many /48s.
>>
>> As this paradigm shift takes place, operators around the world will be
>> clamoring after the RIR for massive blocks, I would say less than /8 more
>> like a /5 or /4.  If you do the math on the way high side a /10 yields 7
>> bits so 128 divided by 5 RIRs yields 24 ISPs per RIR which is tiny number
>> with the number of large size operators worldwide.
>>
>> With the massive proliferation of IOT devices and just about every home
>> or office appliance on 5G, the problem now gets way exacerbated.
>>
>> As this evolution unfolds IANA will be scrambling to release all
>> remaining /3 as well as all unallocated blocks to subvert RIR IPv6 address
>> space depletion.
>>
>> Playing Monday night quarterbacking in hindsight we would never think
>> this would happen in a million years, but we would see IPv6 on the verge of
>> address space depletion.  Unheard of but it can happen as the saying goes
>> “when you build - they will come”.
>>
>> It is true as history has taught that very important lesson.
>>
>> The answer to this real world problem is in the subject heading of this
>> thread.
>>
>> This would also fix the day 1 issue I mentioned allowing mix of slaac
>> devices with static and DHCPv6 up to /80.
>>
>> The variable slaac solution draft proposes a new  RA PIO flag that would
>> be used to signal longer prefixes, and would provide backwards
>> compatibility so that devices not supporting woud ignore the flag  and
>> devices on newer code supporting would use the flag.  We definitely don’t
>> want to impact any existing devices on the existing 64 bit slaac boundary
>> standard.
>>
>> Dmytro has tested the solution on Linux kernel signaling RA PIO flag and
>> was able to successfully test any length mask and random IID generation
>> both RFC 4941 privacy extension as well as stable IID RFC 7217.
>>
>> https://datatracker.ietf.org/doc/draft-mishra-6man-variable-slaac/
>>
>> If everyone is in agreement with what I have stated on this thread, I
>> would like to ask the chairs for WG adoption as this is a WG effort.
>>
>> I would like to garner support from all 6MAN members with full consensus
>> to change the existing RFC 4941 /64 fixed boundary to /80 fixed boundary.
>>
>> Kind Regards
>>
>> Gyan
>>
>> On Thu, Nov 19, 2020 at 5:14 PM Joel M. Halpern <[e-mail]>
>> wrote:
>>
>>> I am missing something in your reasoning.
>>> You seem to say at one point that (to paraphrase) "we can't do this
>>> because it does not work with the existing UE software".
>>> Any new solution where a UE delegates based on any change of any kind
>>> (including lengthening the prefix, shortening the prefix, or magically
>>> incanting new prefixes) requires that the UE be upgraded to know what to
>>> do with the information.  I do not see how that differentiates any of
>>> the solutions. (Except "don't do anything", which I think we do not want
>>> to take.)
>>>
>>> Yours,
>>> Joel
>>>
>>> On 11/19/2020 5:03 PM, Gyan Mishra wrote:
>>> >
>>> >
>>> > On Thu, Nov 19, 2020 at 10:33 AM <[e-mail]
>>> > <mailto:[e-mail]>> wrote:
>>> >
>>> >
>>> >
>>> >      > On 19 Nov 2020, at 14:58, Gyan Mishra <[e-mail]
>>> >     <mailto:[e-mail]>> wrote:
>>> >      >
>>> >      > You would need a new option. It would likely be useful for the
>>> >     requesting router to indicate interest in the option. Even hinting
>>> >     at what prefix size it was expecting.
>>> >      > Now can you explain to me again the reasons why this approach is
>>> >     better than using the existing DHPCv6 protocol packets?
>>> >      >
>>> >      >     3GPP gateway does not support DHCPv6
>>> >
>>> >     3GPP gateway doesn't support new option. What's your point?
>>> >
>>> >
>>> >
>>> >      The point of the v6ops presentation and this email thread is how
>>> to
>>> > “extend a /64” in the 3GPP use case  in slide 1 of the deck you
>>> compiled
>>> > a list of options and of the two I had highlighted in red were the
>>> > 64share v2 Cameron’s option and the variable slaac option.  So on the
>>> > call this morning Lorenzo shot down 64share v2 shorter prefix option
>>> as
>>> > even if the 3GPP architecture was updated to support longer prefixes
>>> and
>>> > even is the 3GPP gateway was able to send a shorter prefix with A flag
>>> > not set, all mobile devices per Lorenzo’s point would be broken as
>>> they
>>> > would not accept the shorter let’s say /56 prefix to build the slaac
>>> 128
>>> > bit address.  So the bottom line is the 64share v2 won’t work unless
>>> we
>>> > update RFC 4291 and remove the 64 bit boundary.
>>> >
>>> >   So we are back to square uno - no viable solution
>>> >
>>> >   So now we had thrown out the longer >64 due to race to bottom
>>> worries
>>> > which I and others believe is Fud and as described in slide 10 of the
>>> > v6ops “race to the bottom slide”.
>>> >
>>> > So a happy medium /80 fixed boundary I came up with that I think
>>> solves
>>> > a lot of the issue and not just the 3GPP initial segmentation of
>>> > downstream devices problem statement.
>>> >
>>> > Since we have to update RFC 4291 for 64share v2 to work anyways to
>>> allow
>>> > for shorter prefixes, why not instead create a new bottom at /80
>>> giving
>>> > 16 bits more of prefix length and shrinking the IID down to 48 bits.
>>> > Doing so you would not even have to update the 3GPP architecture as I
>>> > don’t know if that would fly or not.  Also this solves a few other
>>> > problems at the same time.
>>> >
>>> >
>>> > As I mentioned in the v6ops deck presented that vlsm 0 to 128 is
>>> > mainstream for operators for static addressing on router and switch
>>> > infrastructure and dhcpv6 subnets longer prefixes for network
>>> > infrastructure appliance clusters, NFV/VNF virtualization and server
>>> > farms.  On host subnets where there is a chance of mix of slaac hosts
>>> > with dhcpv6  devices the prefix length is stuck at /64.  So on these
>>> mix
>>> > addressing host subnets we cannot do longer prefixes following our ND
>>> > cache hard limit mantra to prevent ND cache exhaustion issues as
>>> > described in RFC 6164.
>>> >
>>> > So with the /80 new fixed boundary shifting prefix length 16 bits
>>> longer
>>> > and shortening the IID by 16 bits gives resolved the 3GPP issue which
>>> > 64share can work as is and subtending to downstream devices will now
>>> > work as a /64 is now equivalent to a /48 with 64k /80s.  Also BCP-690
>>> > for broadband not all operators have adopted the shorter prefix
>>> lengths
>>> > /56 or /48 recommendations  and now that’s not an issue as the /64
>>> would
>>> > now suffice.
>>> >
>>> >  From an operators perspective that gain allows at least for 3GPP
>>> > massive growth and subtending with a single /64 allows the operators
>>> > such as Verizon with massive subscriber base worldwide can stay with
>>> > current allocations and don’t have to ask for /10.
>>> >
>>> > As 5G gets rolled out with Enhanced VPN framework and Network slicing
>>> > paradigm, the demand for shorter blocks and wearable multiple /48 will
>>> > be our new reality.
>>> >
>>> > Making that 16 bit shift now to /80 making a /64 the new /48 will give
>>> > broadband and 3GPP subscribers a ton of space to subtending their
>>> > networks we would be set for the future.  Especially with IOT the
>>> demand
>>> > for subtending will continue to grow astronomically.
>>> >
>>> > Also IANA does not have to get start in allocating the other /3 and
>>> > other available blocks.
>>> >
>>> > Lots of problems being solved here with a fixed /80 new boundary.
>>> >
>>> > Also with the existing random IID generation schemes which we have
>>> > tested on Linux kernel can do longer p
>>> <https://www.google.com/maps/search/tested+on+Linux+kernel+can+do+longer+p?entry=gmail&source=g>refixes
>>> using RFC 4941 privacy
>>> > extension or RFC 7217 stable IID.
>>> >
>>> > Win-Win for all.
>>> >
>>> >     Ole
>>> >
>>> > --
>>> >
>>> > <http://www.verizon.com/>
>>> >
>>> > *Gyan Mishra*
>>> >
>>> > /Network Solutions A//rchitect /
>>> >
>>> > /M 301 502-1347
>>> > 13101 Columbia Pike
>>> <https://www.google.com/maps/search/13101+Columbia+Pike+%0D%0A+%2FSilver+Spring,+MD?entry=gmail&source=g>
>>>
>>> <https://www.google.com/maps/search/13101+Columbia+Pike+%0D%0A+%2FSilver+Spring,+MD?entry=gmail&source=g>>
>>> /Silver Spring, MD
>>> <https://www.google.com/maps/search/13101+Columbia+Pike+%0D%0A+%2FSilver+Spring,+MD?entry=gmail&source=g>
>>> >
>>> >
>>> >
>>> > _______________________________________________
>>> > v6ops mailing list
>>> > [e-mail]
>>> > https://www.ietf.org/mailman/listinfo/v6ops
>>> >
>>>
>> --
>>
>> <http://www.verizon.com/>
>>
>> *Gyan Mishra*
>>
>> *Network Solutions A**rchitect *
>>
>>
>>
>> *M 301 502-134713101 Columbia Pike
>> <https://www.google.com/maps/search/13101+Columbia+Pike%C2%A0+Silver+Spring,+MD?entry=gmail&source=g>*Silver
>> Spring, MD
>> <https://www.google.com/maps/search/13101+Columbia+Pike%C2%A0+Silver+Spring,+MD?entry=gmail&source=g>
>>
>> --------------------------------------------------------------------
>> IETF IPv6 working group mailing list
>> [e-mail]
>> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
>> --------------------------------------------------------------------
>>
> --

<http://www.verizon.com/>

*Gyan Mishra*

*Network Solutions A**rchitect *



*M 301 502-134713101 Columbia Pike *Silver Spring, MD