Mailing List Archive: nsp: ipv6

Current Consensus on IPv6 Customer Allocation Size

1 2

View All

Index | Next | Previous | View Threaded

tdensmore at tarpit

Aug 1, 2012, 12:17 PM

Post #1 of 44 (1211 views)
Permalink

Current Consensus on IPv6 Customer Allocation Size

Hi Folks,

I have read through multiple threads regarding this issue (though most
of them are years old), and know it may be a can of worms, but I need
some insight into what people are actually doing in 2012. ARIN
"suggests" a /48 for all customers or sites as far as I can tell, though
apparently in the past they also had language including /56 assignments
in some docs. I'm trying to come up with a reasonable numbering plan
that can accommodate /48 customer assignments from our /32.

Basically, here's how I'm looking at things in a nutshell. We currently
have 8 POPs that need subnets allocated, but obviously I want to leave
room for future growth. This leaves me with /36 or /37 per-POP (yes, I
know that the idea of /37 might bother some folks) which would allow me
16 or 32 POPs respectively. Some POPs are obviously smaller than
others, but I don't want to get into variable sized allocations
per-POP. Even with a /36 per-POP, when using /48, this allows me a
maximum of 4096 allocations before having to add a second /36 to the
same POP. This is fine for business connections, but kind of dicey for
residential services. Obviously we could go back to ARIN for another
allocation if we end up in a bind down the road, but there is a real
cost associated with changing designation from a "small" to "large" org
(we actually qualify as a medium org, but nibble boundary allocations)
that I'd prefer to avoid.

Is the current (again, 2012 - most threads and books that I have read
are al least a few years old) consensus that a /48 per-residential-user
really justified? Opinions or pointers to current Fine Manuals to read
would be most appreciated.

Thanks,

TD

chip.gwyn at gmail

Aug 1, 2012, 12:41 PM

Post #2 of 44 (1163 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

Hi Tim,

So far we're doing the following:

- /32 per metro area or region
- break that into /36's for each pop/data-center
- then supply customers with /56's by default
- if they want more or want to multi-home, hand them a /48
- Point to point interfaces within the network = /126's
- customer facing interfaces are /112's or /64's depending on the case or needs

--chip

On Wed, Aug 1, 2012 at 3:17 PM, Tim Densmore
<tdensmore [at] tarpit> wrote:
> Hi Folks,
>
> I have read through multiple threads regarding this issue (though most of
> them are years old), and know it may be a can of worms, but I need some
> insight into what people are actually doing in 2012. ARIN "suggests" a /48
> for all customers or sites as far as I can tell, though apparently in the
> past they also had language including /56 assignments in some docs. I'm
> trying to come up with a reasonable numbering plan that can accommodate /48
> customer assignments from our /32.
>
> Basically, here's how I'm looking at things in a nutshell. We currently
> have 8 POPs that need subnets allocated, but obviously I want to leave room
> for future growth. This leaves me with /36 or /37 per-POP (yes, I know that
> the idea of /37 might bother some folks) which would allow me 16 or 32 POPs
> respectively. Some POPs are obviously smaller than others, but I don't want
> to get into variable sized allocations per-POP. Even with a /36 per-POP,
> when using /48, this allows me a maximum of 4096 allocations before having
> to add a second /36 to the same POP. This is fine for business connections,
> but kind of dicey for residential services. Obviously we could go back to
> ARIN for another allocation if we end up in a bind down the road, but there
> is a real cost associated with changing designation from a "small" to
> "large" org (we actually qualify as a medium org, but nibble boundary
> allocations) that I'd prefer to avoid.
>
> Is the current (again, 2012 - most threads and books that I have read are al
> least a few years old) consensus that a /48 per-residential-user really
> justified? Opinions or pointers to current Fine Manuals to read would be
> most appreciated.
>
> Thanks,
>
> TD

--
Just my $.02, your mileage may vary, batteries not included, etc....

cgrundemann at gmail

Aug 1, 2012, 12:49 PM

Post #3 of 44 (1161 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

Best Current Operational Practice (BCOP) on IPv6 Subnetting:
http://www.ipbcop.org/ratified-bcops/bcop-ipv6-subnetting/

/48 per site is best. I would highly recommend swallowing the ~$2k/yr
and get the allocation you need now, so that your network can grow in
a structured, homogenous manner. Rather than fighting fires later to
save a buck now (I mean, I have to guess that buying even one router a
year blows that cost out of the water anyway - even a line card...).

Cheers,
~Chris

On Wed, Aug 1, 2012 at 1:17 PM, Tim Densmore
<tdensmore [at] tarpit> wrote:
> Hi Folks,
>
> I have read through multiple threads regarding this issue (though most of
> them are years old), and know it may be a can of worms, but I need some
> insight into what people are actually doing in 2012. ARIN "suggests" a /48
> for all customers or sites as far as I can tell, though apparently in the
> past they also had language including /56 assignments in some docs. I'm
> trying to come up with a reasonable numbering plan that can accommodate /48
> customer assignments from our /32.
>
> Basically, here's how I'm looking at things in a nutshell. We currently
> have 8 POPs that need subnets allocated, but obviously I want to leave room
> for future growth. This leaves me with /36 or /37 per-POP (yes, I know that
> the idea of /37 might bother some folks) which would allow me 16 or 32 POPs
> respectively. Some POPs are obviously smaller than others, but I don't want
> to get into variable sized allocations per-POP. Even with a /36 per-POP,
> when using /48, this allows me a maximum of 4096 allocations before having
> to add a second /36 to the same POP. This is fine for business connections,
> but kind of dicey for residential services. Obviously we could go back to
> ARIN for another allocation if we end up in a bind down the road, but there
> is a real cost associated with changing designation from a "small" to
> "large" org (we actually qualify as a medium org, but nibble boundary
> allocations) that I'd prefer to avoid.
>
> Is the current (again, 2012 - most threads and books that I have read are al
> least a few years old) consensus that a /48 per-residential-user really
> justified? Opinions or pointers to current Fine Manuals to read would be
> most appreciated.
>
> Thanks,
>
> TD

--
@ChrisGrundemann
http://chrisgrundemann.com

marc.blanchet at viagenie

Aug 1, 2012, 1:29 PM

Post #4 of 44 (1165 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

Consider reading RFC3531 for allocation methods that could help you have more flexibility in your address space, whatever initial boundaries you decide at the beginning.

Marc.

Le 2012-08-01 � 12:17, Tim Densmore a �crit :

> Hi Folks,
>
> I have read through multiple threads regarding this issue (though most of them are years old), and know it may be a can of worms, but I need some insight into what people are actually doing in 2012. ARIN "suggests" a /48 for all customers or sites as far as I can tell, though apparently in the past they also had language including /56 assignments in some docs. I'm trying to come up with a reasonable numbering plan that can accommodate /48 customer assignments from our /32.
>
> Basically, here's how I'm looking at things in a nutshell. We currently have 8 POPs that need subnets allocated, but obviously I want to leave room for future growth. This leaves me with /36 or /37 per-POP (yes, I know that the idea of /37 might bother some folks) which would allow me 16 or 32 POPs respectively. Some POPs are obviously smaller than others, but I don't want to get into variable sized allocations per-POP. Even with a /36 per-POP, when using /48, this allows me a maximum of 4096 allocations before having to add a second /36 to the same POP. This is fine for business connections, but kind of dicey for residential services. Obviously we could go back to ARIN for another allocation if we end up in a bind down the road, but there is a real cost associated with changing designation from a "small" to "large" org (we actually qualify as a medium org, but nibble boundary allocations) that I'd prefer to avoid.
>
> Is the current (again, 2012 - most threads and books that I have read are al least a few years old) consensus that a /48 per-residential-user really justified? Opinions or pointers to current Fine Manuals to read would be most appreciated.
>
> Thanks,
>
> TD

tdensmore at tarpit

Aug 1, 2012, 2:03 PM

Post #5 of 44 (1160 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

Hi Chris,

Thanks for the link! I had actually read through it recently and found
it quite useful. I find the back and forth between /64 vs /12[67] on
P2P links interesting.

I've read rfc 3177, and though some of it was above my head, I couldn't
really find anything in it that really defined *why* a /48 was preferred
(GSE, 6to4, and Site Local require /48, so /48 is best was what I took
away). As far as I can tell, the layman arguments in favor go like this:

- /48 is likely to be equal to /24 (v4) bgp-wise, and anything longer is
likely to be filtered.

- /48 for *everyone* allows for uniform customer allocation size.

- It's "easy" to forsee that someday in the future people will need more
than 256 subnets in the home, and since nibble boundaries are considered
a must, then /48 is the only option.

Is that about right?

I think I'd have a hard time pitching extra cash outlay for a /28.

Thanks again!

TD

On 8/1/2012 1:49 PM, Chris Grundemann wrote:
> Best Current Operational Practice (BCOP) on IPv6 Subnetting:
> http://www.ipbcop.org/ratified-bcops/bcop-ipv6-subnetting/
>
> /48 per site is best. I would highly recommend swallowing the ~$2k/yr
> and get the allocation you need now, so that your network can grow in
> a structured, homogenous manner. Rather than fighting fires later to
> save a buck now (I mean, I have to guess that buying even one router a
> year blows that cost out of the water anyway - even a line card...).
>
> Cheers,
> ~Chris

mark at exonetric

Aug 1, 2012, 2:12 PM

Post #6 of 44 (1170 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On 1 Aug 2012, at 22:03, Tim Densmore <tdensmore [at] tarpit> wrote:
>
>
> - It's "easy" to forsee that someday in the future people will need more than 256 subnets in the home, and since nibble boundaries are considered a must, then /48 is the only option.

More than 256 subnets in the home? Who would want to manage all of that?

I can't see more than about 16 subnets at a single residential
address with a single family and would suggest a /60 by default
for "bog standard" residential customers, with a /56 possible
if they ask nicely. Every subnet represents a management burden
for the end user, if a genuinely separate subnet is required.

- Mark

tdensmore at tarpit

Aug 1, 2012, 2:20 PM

Post #7 of 44 (1157 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On 8/1/2012 3:12 PM, Mark Blackman wrote:
> More than 256 subnets in the home? Who would want to manage all of that?

Hi Mark,

Thanks for the input! It's not an argument I'm making, but one I have
seen made. Something along the lines of "in the future your fridge and
TV will each need their own own subnet" - that kind of thought.
Obviously that'd be a ways down the road.

Thanks,

TD

tdensmore at tarpit

Aug 1, 2012, 2:22 PM

Post #8 of 44 (1166 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On 8/1/2012 1:41 PM, chip wrote:
> Hi Tim,
>
> So far we're doing the following:
>
> - /32 per metro area or region
> - break that into /36's for each pop/data-center
> - then supply customers with /56's by default
> - if they want more or want to multi-home, hand them a /48
> - Point to point interfaces within the network = /126's
> - customer facing interfaces are /112's or /64's depending on the case or needs
>
> --chip

Hi Chip,

Thanks for the reply! This is exactly the sort of info I was after -
what's actually being done versus what's indicated in a book from 2006.

TD

cgrundemann at gmail

Aug 1, 2012, 2:24 PM

Post #9 of 44 (1161 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On Wed, Aug 1, 2012 at 3:03 PM, Tim Densmore
<tdensmore [at] tarpit> wrote:
> Hi Chris,
>
> Thanks for the link! I had actually read through it recently and found it
> quite useful. I find the back and forth between /64 vs /12[67] on P2P links
> interesting.
>
> I've read rfc 3177, and though some of it was above my head, I couldn't
> really find anything in it that really defined *why* a /48 was preferred
> (GSE, 6to4, and Site Local require /48, so /48 is best was what I took
> away). As far as I can tell, the layman arguments in favor go like this:
>
> - /48 is likely to be equal to /24 (v4) bgp-wise, and anything longer is
> likely to be filtered.

True.

> - /48 for *everyone* allows for uniform customer allocation size.

Right again.

> - It's "easy" to forsee that someday in the future people will need more
> than 256 subnets in the home, and since nibble boundaries are considered a
> must, then /48 is the only option.

Maybe. What is more likely (imo) is that we will see more and more
complex self-configuring home networks. The /48 gives you 16 bits to
play with which makes algorithmic subnetting without human
intervention much more possible. I don't know that I'd say we will
definitely need >256 subnets, rather that we need the flexibility to
"plug and play" routers in a multi-tier home network. If we break a
home network down by "width" and "depth" it looks something like this:

ISP
|
1 CER [/48] (Customer Edge Router)
|
15 L1IRs [/52] (Level 1 Internal Routers)
|
15 L2IRs [/56]
|
15 L3IRs [/60]
|
15 L4IRs [/64]

If you use 16 bits for internal subnetting (a /48) you get upto 15
routers (wide) at every level and upto 4 levels of routers (deep)
within the network. You also know that this is the case
deterministically. This makes prefix autoconfiguration easy (if I get
a /56 I'm a L2IR and I should hand out /60s downstream).

You may not go to full width at each level, nor to full depth in each
branch, but having those 16 bits gives you a stable framework to
operate in. Yes, you can get away with less - but why? The future of
innovation is brighter the more flexibility we provide today, and IPv6
is abundant (not scarce like IPv4)[1].

> Is that about right?
>
> I think I'd have a hard time pitching extra cash outlay for a /28.
>
> Thanks again!

Cheers,
~Chris

> TD
>
>
>
> On 8/1/2012 1:49 PM, Chris Grundemann wrote:
>>
>> Best Current Operational Practice (BCOP) on IPv6 Subnetting:
>> http://www.ipbcop.org/ratified-bcops/bcop-ipv6-subnetting/
>>
>> /48 per site is best. I would highly recommend swallowing the ~$2k/yr
>> and get the allocation you need now, so that your network can grow in
>> a structured, homogenous manner. Rather than fighting fires later to
>> save a buck now (I mean, I have to guess that buying even one router a
>> year blows that cost out of the water anyway - even a line card...).
>>
>> Cheers,
>> ~Chris
>
>

--
@ChrisGrundemann
http://chrisgrundemann.com

eugen at leitl

Aug 1, 2012, 2:25 PM

Post #10 of 44 (1162 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On Wed, Aug 01, 2012 at 03:20:58PM -0600, Tim Densmore wrote:
> On 8/1/2012 3:12 PM, Mark Blackman wrote:
>> More than 256 subnets in the home? Who would want to manage all of that?
>
> Hi Mark,
>
> Thanks for the input! It's not an argument I'm making, but one I have
> seen made. Something along the lines of "in the future your fridge and
> TV will each need their own own subnet" - that kind of thought.
> Obviously that'd be a ways down the road.

In IPv6 land, your /64 is your /32 in IPv4. Given that some of
us run out of IPv4 /24 at home (nevermind the virtual hosts,
kilonode and meganode hardware is coming) a /48 for each
residential user appears a no-brainer.

mark at exonetric

Aug 1, 2012, 2:35 PM

Post #11 of 44 (1162 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On 1 Aug 2012, at 22:25, Eugen Leitl <eugen [at] leitl> wrote:

> On Wed, Aug 01, 2012 at 03:20:58PM -0600, Tim Densmore wrote:
>> On 8/1/2012 3:12 PM, Mark Blackman wrote:
>>> More than 256 subnets in the home? Who would want to manage all of that?
>>
>> Hi Mark,
>>
>> Thanks for the input! It's not an argument I'm making, but one I have
>> seen made. Something along the lines of "in the future your fridge and
>> TV will each need their own own subnet" - that kind of thought.
>> Obviously that'd be a ways down the road.
>
> In IPv6 land, your /64 is your /32 in IPv4. Given that some of
> us run out of IPv4 /24 at home (nevermind the virtual hosts,
> kilonode and meganode hardware is coming) a /48 for each
> residential user appears a no-brainer.

I'm arguing that's the uncommon case and operators should
have a default prefix that's closer to the common case, but
have some mechanism for allocating to self-described power-users.
I could just about see a /56 by default, but not ever a /48 by
default for a single family dwelling.

A genuinely separate subnet is a management burden, which
I'd guess that 95% or more of single family residential
customers simply aren't looking for.

- Mark

eugen at leitl

Aug 1, 2012, 2:42 PM

Post #12 of 44 (1162 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On Wed, Aug 01, 2012 at 10:35:45PM +0100, Mark Blackman wrote:

> > In IPv6 land, your /64 is your /32 in IPv4. Given that some of
> > us run out of IPv4 /24 at home (nevermind the virtual hosts,
> > kilonode and meganode hardware is coming) a /48 for each
> > residential user appears a no-brainer.
>
> I'm arguing that's the uncommon case and operators should

Currently, absolutely. However, in future there tends to
be a feature creep, so I would overengineer today so that
I can last longer tomorrow.

> have a default prefix that's closer to the common case, but
> have some mechanism for allocating to self-described power-users.
> I could just about see a /56 by default, but not ever a /48 by
> default for a single family dwelling.

How much additional overhead is this? IPv6 isn't forever,
anyway. It would be prudent it lasts about as long as IPv4
did.

> A genuinely separate subnet is a management burden, which

The interesting part is complete automation. The user
wouldn't have the foggiest, it's all just automagick.

But I'd rather think of /64 for each device, not just
LAN segment. So /48 for each home makes sense.

> I'd guess that 95% or more of single family residential
> customers simply aren't looking for.

Ideally, they would have no clue what this IPv4/IPv6 thing
is all about.

dr at cluenet

Aug 1, 2012, 3:13 PM

Post #13 of 44 (1160 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On Wed, Aug 01, 2012 at 10:12:55PM +0100, Mark Blackman wrote:
> More than 256 subnets in the home? Who would want to manage all of that?
>
> I can't see more than about 16 subnets at a single residential
> address with a single family

"I think there is a world market for maybe five computers."
- Thomas Watson, Chairman, IBM, 1943

"There is no reason why anyone would want a computer in their home."
- Ken Olsen, Founder, DEC, 1977

"640K should keep everyone happy for the next 10 years"
- Bill Gates, Founder, Microsoft, 1981

"4 bits for home subnetting should be more than enough"
- any takers?

Best regards, SCNR,
Daniel

--
CLUE-RIPE -- Jabber: dr [at] cluenet -- dr [at] IRCne -- PGP: 0xA85C8AA0

sm at resistor

Aug 1, 2012, 3:22 PM

Post #14 of 44 (1160 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

At 12:17 PM 8/1/2012, Tim Densmore wrote:
>Is the current (again, 2012 - most threads and books that I have
>read are al least a few years old) consensus that a /48
>per-residential-user really justified? Opinions or pointers to
>current Fine Manuals to read would be most appreciated.

RFC 6177 clarifies the /48 recommendation. The Fine Manuals won't
really explain why it has been argued that residential users should
be given a /56.

Justification is appropriate if you are dealing with a scare
resource. There is a management overhead in dealing with that. It
is easier to hand out ample address space so that people can do what
they want without having to come back to you for additional address
space. A /48 keeps it simple.

Regards,
-sm

dougb at dougbarton

Aug 1, 2012, 3:34 PM

Post #15 of 44 (1163 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On 8/1/2012 2:12 PM, Mark Blackman wrote:
>
> On 1 Aug 2012, at 22:03, Tim Densmore <tdensmore [at] tarpit> wrote:
>>
>>
>> - It's "easy" to forsee that someday in the future people will need more than 256 subnets in the home, and since nibble boundaries are considered a must, then /48 is the only option.
>
> More than 256 subnets in the home? Who would want to manage all of that?

As Chris pointed out there is a theoretical (not unlikely, just not here
yet) future where we will have massive, smart, cascading home networks.
So, since v6 addresses are not scarce, why not allow for this to happen
in the initial phases of planning?

However, as you correctly point out Mark, hand a user a /48 and they're
lost.

So, the plan I have been advocating for years now is to do what the RIRs
are doing. Start your addressing plan by *reserving* the largest block
you think may be reasonable for each customer, but today you only assign
something more manageable. That way you can always go back and let a
power user expand into their reservation without disrupting anything.
And if down the road you end up with reservations that you're confident
are excessive, you can divide them and start the reservation/assignment
process over.

So in Tim's case I'd reserve a /48, and assign the first /56 from it.
Personally I think a /60 is more than enough for a home user nowadays,
but IIRC ARIN's guidance on utilization mentions a /56. (You're on your
own to research/confirm that though.) :)

And Tim, one other factor you alluded to that I think it's worth making
explicit. You want to keep the assignments on the nibble boundaries
because it makes DNS management easier.

hth,

Doug

--
If you're never wrong, you're not trying hard enough

tony at lavanauts

Aug 1, 2012, 9:01 PM

Post #16 of 44 (1160 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On Wed, 1 Aug 2012, Mark Blackman wrote:

> More than 256 subnets in the home? Who would want to manage all of that?

> if they ask nicely. Every subnet represents a management burden
> for the end user, if a genuinely separate subnet is required.

The reality is that customers buy multiple off-the-shelf routers and
expect them to work without intervention. In the IPv6 world that means PD
needs to work automatically and recursively among the chained routers.
Depending on how these devices are nested, and how they do PD, the poor
router at the end of a long chain may not have a whole lot of /64s to play
with :)

--
Antonio Querubin
e-mail: tony [at] lavanauts
xmpp: antonioquerubin [at] gmail

dougb at dougbarton

Aug 1, 2012, 9:13 PM

Post #17 of 44 (1160 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On 8/1/2012 9:01 PM, Antonio Querubin wrote:
> The reality is that customers buy multiple off-the-shelf routers

Evidence please.

I think (and no, I don't have hard facts here either, just experience)
that the vast majority of home users have 1 network device, and could
live quite happily with 1 v6 subnet.

The people who buy, and use, multiple routers at the same time are us. :)

Doug

--

I am only one, but I am one. I cannot do everything, but I can do
something. And I will not let what I cannot do interfere with what
I can do.
-- Edward Everett Hale, (1822 - 1909)

tony at lavanauts

Aug 1, 2012, 9:40 PM

Post #18 of 44 (1161 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On Wed, 1 Aug 2012, Doug Barton wrote:

> Evidence please.
>
> I think (and no, I don't have hard facts here either, just experience)
> that the vast majority of home users have 1 network device, and could
> live quite happily with 1 v6 subnet.
>
> The people who buy, and use, multiple routers at the same time are us. :)

Did I mention home user? :)

But even so, a quick, non-scientific survey of what my friends run 'at
home' suggests about 1 in 4 run more than one routing device in the home
if for no reason than to get either additional wired ports or extend WIFI
coverage.

Looking at some SOHO type businesses I've dealt with, a majority of them
do have multiple routing devices. Again not just because they need them
to actually run multiple subnets but because the business grew and they
needed more ports in more places. Regardless the devices just have to
work. Multiply NAT'ed IPv4 will be replaced by multiply nested IPv6 PD.

--
Antonio Querubin
e-mail: tony [at] lavanauts
xmpp: antonioquerubin [at] gmail

dougb at dougbarton

Aug 1, 2012, 9:48 PM

Post #19 of 44 (1161 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On 8/1/2012 9:40 PM, Antonio Querubin wrote:
> On Wed, 1 Aug 2012, Doug Barton wrote:
>
>> Evidence please.
>>
>> I think (and no, I don't have hard facts here either, just experience)
>> that the vast majority of home users have 1 network device, and could
>> live quite happily with 1 v6 subnet.
>>
>> The people who buy, and use, multiple routers at the same time are us. :)
>
> Did I mention home user? :)

Fair enough.

> But even so, a quick, non-scientific survey of what my friends

Your friends (by definition) are outliers on the bell-shaped curve.

> run 'at
> home' suggests about 1 in 4 run more than one routing device in the home
> if for no reason than to get either additional wired ports or extend
> WIFI coverage.

Extending wifi coverage doesn't need cascading PD, it needs a bridge.

> Looking at some SOHO type businesses I've dealt with, a majority of them
> do have multiple routing devices. Again not just because they need them
> to actually run multiple subnets but because the business grew and they
> needed more ports in more places. Regardless the devices just have to
> work.

My experience of the small to midsize business market is similar to
yours, and I think "just work" is a nice goal, but that also doesn't
necessarily mean cascading PD. In fact, I think that to the extent they
would notice at all most SOHO users would be quite surprised if the fact
that something was plugged into one jack or another meant that it
received a totally different address.

Just because with v6 you _can_ have lots of networks, cascading or
otherwise, doesn't mean that you should.

> Multiply NAT'ed IPv4 will be replaced by multiply nested IPv6 PD.

I sincerely hope not.

--

I am only one, but I am one. I cannot do everything, but I can do
something. And I will not let what I cannot do interfere with what
I can do.
-- Edward Everett Hale, (1822 - 1909)

seth.mos at dds

Aug 1, 2012, 10:38 PM

Post #20 of 44 (1165 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

Hi,

Op 1 aug 2012, om 23:12 heeft Mark Blackman het volgende geschreven:

>
> On 1 Aug 2012, at 22:03, Tim Densmore <tdensmore [at] tarpit> wrote:
>>
>>
>> - It's "easy" to forsee that someday in the future people will need more than 256 subnets in the home, and since nibble boundaries are considered a must, then /48 is the only option.
>
> More than 256 subnets in the home? Who would want to manage all of that?

You're approaching this the wrong way. Try from a computer/router standpoint, not human.

So you have the initial router that gets a /48 assignment (for example through DHCP-PD), this router then hands out DHCP-PD as well, but /56 in size. But you need to allocate part of the existing /48 for this pool which you then can not use for anything else.

You don't want to hand out just a /64 because you don't know what's behind it.

It means that even if they buy *a* router at your favourite electronics store and hook it up to the existing router (because they wanted better WIFI) or something else. They still would have native IPv6 internet without hoops and it would just work. Atleast a bit more intuitive then it does now.

Disclaimer: I am a developer of the open source pfSense firewall. It does just this, I've tested upto 3 levels deep.

Assigning a /60 pretty makes any sort of flexible approach to this impossible, it's too bad that the previous experience has made us apply the same reasoning to a plenty large address space.

Regards,

Seth

dwcarder at wisc

Aug 1, 2012, 10:52 PM

Post #21 of 44 (1159 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On Aug 1, 2012, at 11:13 PM, Doug Barton wrote:

> On 8/1/2012 9:01 PM, Antonio Querubin wrote:
>> The reality is that customers buy multiple off-the-shelf routers
>
> Evidence please.
>
> I think (and no, I don't have hard facts here either, just experience)
> that the vast majority of home users have 1 network device, and could
> live quite happily with 1 v6 subnet.

At least at this point in time, one single L2 domain is
essentially required for automagic device discovery stuff
like mDNS to work among consumer devices.

Dale

roger at jorgensen

Aug 1, 2012, 11:28 PM

Post #22 of 44 (1153 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

> On Wed, Aug 01, 2012 at 10:12:55PM +0100, Mark Blackman wrote:
>> More than 256 subnets in the home? Who would want to manage all of that?
>>
>> I can't see more than about 16 subnets at a single residential
>> address with a single family
>
> "I think there is a world market for maybe five computers."
> - Thomas Watson, Chairman, IBM, 1943
>
> "There is no reason why anyone would want a computer in their home."
> - Ken Olsen, Founder, DEC, 1977
>
> "640K should keep everyone happy for the next 10 years"
> - Bill Gates, Founder, Microsoft, 1981
>
> "4 bits for home subnetting should be more than enough"
> - any takers?

People keep forgetting those famous "last" words, but more important, we
really need to stop thinking in the old terms about our homenetwork. We're
not supposed to manage 256 subnets, they are part of our homenetwork that
should just work.
Plenty of working group and other discussion going on about that subject,
homenet [at] iet is one. Go there and start looking into all of the different
issues they're addressing, there are some of them...

--
------------------------------
Roger Jorgensen | - ROJO9-RIPE - RJ1866P-NORID
roger [at] jorgensen | - The Future is IPv6
-------------------------------------------------------

A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

swmike at swm

Aug 1, 2012, 11:30 PM

Post #23 of 44 (1153 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On Wed, 1 Aug 2012, Tim Densmore wrote:

> trying to come up with a reasonable numbering plan that can accommodate /48
> customer assignments from our /32.

First of all, /32 is the default allocation you get without justification,
right. So if you justify, you will get more. Do a proper justification
now, not in 3 years, so you're on the right track by trying to find out.

If you want a big subnet, planning to hand out /48 to everybody will give
you a lot of addresses, increasing flexibility down the road.

Personally I advocate handing out /56 to residential and /48 to business.
This just seems like a reasonable compromise between "should be plenty for
99% or more of the use cases" and let's aim for the sky. /60 seems short
sighted for residential.

--
Mikael Abrahamsson email: swmike [at] swm

dr at cluenet

Aug 2, 2012, 12:49 AM

Post #24 of 44 (1158 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On Thu, Aug 02, 2012 at 12:52:22AM -0500, Dale W. Carder wrote:
> At least at this point in time, one single L2 domain is
> essentially required for automagic device discovery stuff
> like mDNS to work among consumer devices.

As far as I understood, smart metering networks e.g. need a routing
device as the underlying technology can't be bridged to Ethernet.
Also, they are thinking about multihomed home networks which e.g. helps
migrating between providers, or for redundancy.

No kidding, but they (IETF) are actually sporting the idea of running
OSPF in residential LANs to accomodate such scenarios in a plug-and-play
manner. :-)

Best regards,
Daniel

--
CLUE-RIPE -- Jabber: dr [at] cluenet -- dr [at] IRCne -- PGP: 0xA85C8AA0

mark at exonetric

Aug 2, 2012, 2:07 AM

Post #25 of 44 (1160 views)
Permalink

Re: Current Consensus on IPv6 Customer Allocation Size [In reply to]

On 1 Aug 2012, at 23:13, Daniel Roesen <dr [at] cluenet> wrote:

> On Wed, Aug 01, 2012 at 10:12:55PM +0100, Mark Blackman wrote:
>> More than 256 subnets in the home? Who would want to manage all of that?
>>
>> I can't see more than about 16 subnets at a single residential
>> address with a single family
>
> "I think there is a world market for maybe five computers."
> - Thomas Watson, Chairman, IBM, 1943
>
> "There is no reason why anyone would want a computer in their home."
> - Ken Olsen, Founder, DEC, 1977
>
> "640K should keep everyone happy for the next 10 years"
> - Bill Gates, Founder, Microsoft, 1981
>
> "4 bits for home subnetting should be more than enough"
> - any takers?

Comcast alone has around 2^26 customers (45.2 million)

"24.4 million cable customers, 14.7 million high-speed Internet
customers and 6.1 million digital voice customers in 39 states and the
District of Columbia." http://business.comcast.com/smb/about

Comcast's only allocation that I can see is a /31,
http://whois.arin.net/rest/net/NET6-2001-558-1.html

They're going to need a much bigger IPv6 allocation to give all ~2^26
customers a /48 allocation, and they can't even quite give them all a
/56 out of their current allocation.

Did they get their initial allocation wrong? Should all 45 million get a
/48 ?

I'd guess that these numbers are much bigger by one or two orders of
decimal magnitude in Asia.

- Mark

1 2

View All

Index | Next | Previous | View Threaded

Interested in having your list archived? Contact Gossamer Threads