rfc9663.original.xml   rfc9663.xml 
<?xml version="1.0" encoding="utf-8"?> <?xml version="1.0" encoding="UTF-8"?>
<?xml-model href="rfc7991bis.rnc"?> <!-- Required for schema validation and sch
ema-aware editing --> <!-- draft submitted in xml v3 -->
<!-- <?xml-stylesheet type="text/xsl" href="rfc2629.xslt" ?> -->
<!-- This third-party XSLT can be enabled for direct transformations in XML proc
essors, including most browsers -->
<!DOCTYPE rfc [ <!DOCTYPE rfc [
<!ENTITY nbsp "&#160;"> <!ENTITY nbsp "&#160;">
<!ENTITY zwsp "&#8203;"> <!ENTITY zwsp "&#8203;">
<!ENTITY nbhy "&#8209;"> <!ENTITY nbhy "&#8209;">
<!ENTITY wj "&#8288;"> <!ENTITY wj "&#8288;">
]> ]>
<!-- If further character entities are required then they should be added to the
DOCTYPE above.
Use of an external entity file is not recommended. -->
<rfc <rfc xmlns:xi="http://www.w3.org/2001/XInclude" category="info" docName="draft-i
xmlns:xi="http://www.w3.org/2001/XInclude" etf-v6ops-dhcp-pd-per-device-08" number="9663" ipr="trust200902" obsoletes="" up
category="info" dates="" submissionType="IETF" consensus="true" xml:lang="en" version="3" sortRe
docName="draft-ietf-v6ops-dhcp-pd-per-device-08" fs="true" symRefs="true">
ipr="trust200902"
obsoletes=""
updates=""
submissionType="IETF"
xml:lang="en"
version="3">
<front> <front>
<title abbrev="Prefix per client using DHCPv6 PD">Using DHCPv6-PD to Al <title abbrev="Prefix per Client Using DHCPv6-PD">Using DHCPv6 Prefix
locate Unique IPv6 Prefix per Client in Large Broadcast Networks</title> Delegation (DHCPv6-PD) to Allocate Unique IPv6 Prefixes per Client in Large
Broadcast Networks</title>
<seriesInfo name="Internet-Draft" value="draft-ietf-v6ops-dhcp-pd-per-d <seriesInfo name="RFC" value="9663"/>
evice"/>
<author initials="L." surname="Colitti" fullname="Lorenzo Colitti"> <author initials="L." surname="Colitti" fullname="Lorenzo Colitti">
<organization>Google, LLC</organization> <organization>Google, LLC</organization>
<address> <address>
<postal> <postal>
<street>Shibuya 3-21-3</street> <street>Shibuya 3-21-3</street>
<country>Japan</country> <country>Japan</country>
</postal> </postal>
<email>lorenzo@google.com</email> <email>lorenzo@google.com</email>
</address> </address>
</author> </author>
<author fullname="Jen Linkova" initials="J" role="editor" surname="Linkova"> <author fullname="Jen Linkova" initials="J" role="editor" surname="Linkova">
<organization>Google</organization> <organization>Google</organization>
<address> <address>
<postal> <postal>
<!-- Reorder these if your country does things differently -->
<street>1 Darling Island Rd</street> <street>1 Darling Island Rd</street>
<city>Pyrmont</city> <city>Pyrmont</city>
<region>NSW</region> <region>New South Wales</region>
<code>2009</code> <code>2009</code>
<country>AU</country> <country>Australia</country>
</postal> </postal>
<email>furry13@gmail.com</email> <email>furry13@gmail.com</email>
<email>furry@google.com</email> <email>furry@google.com</email>
</address> </address>
</author> </author>
<author fullname="Xiao Ma" initials="X" role="editor" surname="Ma"> <author fullname="Xiao Ma" initials="X" role="editor" surname="Ma">
<organization>Google</organization> <organization>Google</organization>
<address> <address>
<postal> <postal>
<street>Shibuya 3-21-3</street> <street>Shibuya 3-21-3</street>
<country>Japan</country> <country>Japan</country>
</postal> </postal>
<email>xiaom@google.com</email> <email>xiaom@google.com</email>
</address> </address>
</author> </author>
<date month="September" year="2024"/>
<area>OPS</area>
<workgroup>v6ops</workgroup>
<date year="2024"/>
<area>OPS Area</area>
<workgroup>v6ops Working Group</workgroup>
<keyword>IPv6</keyword> <keyword>IPv6</keyword>
<keyword>SLAAC</keyword> <keyword>SLAAC</keyword>
<keyword>DHCPv6-PD</keyword> <keyword>DHCPv6-PD</keyword>
<abstract> <abstract>
<t>This document discusses an IPv6 deployment scenario when individua <t>This document discusses an IPv6 deployment scenario when individual
l nodes connected to large broadcast networks (such as enterprise networks or pu nodes connected to large broadcast networks (such as enterprise networks
blic Wi-Fi networks) are allocated unique prefixes via DHCPv6 Prefix Delegation or public Wi-Fi networks) are allocated unique prefixes via DHCPv6
(DHCPv6-PD, RFC8415).</t> Prefix Delegation (DHCPv6-PD), as specified in RFC 8415.</t>
</abstract> </abstract>
</front> </front>
<middle> <middle>
<section> <section>
<name>Introduction</name> <name>Introduction</name>
<t>Often broadcast networks such as enterprise or public Wi-Fi deployments place many devices on a shared link with a single on-link prefix. This document describes an alternative deployment model where individual devices obtain prefi xes from the network. This provides two important advantages.</t>
<t>First, it offers better scalability. Unlike IPv4, IPv6 allows hosts to <t>Often, broadcast networks such as enterprise or public Wi-Fi
have multiple addresses, and this is the case in most deployments (see <xref tar deployments place many devices on a shared link with a single on-link
get="appendix"/> for more details). prefix. This document describes an alternative deployment model where
individual devices obtain prefixes from the network. This provides two
important advantages.</t>
However, increasing the number of addresses introduces scalability issues on the <!--[rfced] May we rework the text below for clarity and concision?
network infrastructure.
Network devices need to maintain various types of tables/hashes (Neighbor Cache
on first-hop routers, Neighbor Discovery Proxy caches on Layer 2 devices etc.).
On VXLAN <xref target="RFC7348"/> networks each address mig
ht be represented as a route so 8 addresses instead of 1 requires the devices to
support 8 times more routes, etc.
If an infrastructure device resources are exhausted, the de
vice might drop some IPv6 addresses from the corresponding tables, while the add
ress owner might be still using the address to send traffic. This leads to traff
ic blackholing and degraded customer experience.
Providing every host with one prefix allows the network to maintain only one ent
ry per device, while still providing the device the ability to use an arbitrary
number of addresses.
</t>
<t>Second, it provides the ability to extend the network. In IPv4, a devic e that connects to the network can provide connectivity to subtended devices by using NAT. With DHCPv6 Prefix Delegation (DHCPv6-PD, <xref target="RFC8415"/>), such a device can similarly extend the network, but unlike IPv4 NAT, it can prov ide its subtended devices with full end-to-end connectivity.</t> Original:
<t>Another method of deploying unique prefixes per device is documented in On VXLAN [RFC7348] networks each address might be represented as a route so
<xref target="RFC8273"/>. Similarly, the standard deployment model in cellular 8 addresses instead of 1 requires the devices to support 8 times more
IPv6 networks <xref target="RFC6459"/> provides a unique prefix to every device. routes, etc.
However, providing a unique prefix per device is very uncommon in enterprise-st
yle networks, where nodes are usually connected to broadcast segments/VLANs and Perhaps:
each link has a single on-link prefix assigned. This document takes a similar ap
proach to <xref target="RFC8273"/>, but allocates the prefix using DHCPv6-PD. On Virtual eXtensible Local Area Network (VXLAN) networks [RFC7348], each
</t> address might be represented as a route; this means that eight addresses,
<t>This document focuses on the behaviour of the network. Host behaviour i instead of one, require the devices to support eight times more routes, etc.
s not defined in this document. -->
<t>First, it offers better scalability. Unlike IPv4, IPv6 allows hosts
to have multiple addresses, and this is the case in most deployments
(see <xref target="appendix"/> for more details). However, increasing
the number of addresses introduces scalability issues on the network
infrastructure. Network devices need to maintain various types of
tables/hashes (Neighbor Cache on first-hop routers, Neighbor Discovery
Proxy caches on Layer 2 devices, etc.). On Virtual eXtensible Local Area
Network (VXLAN) networks <xref
target="RFC7348"/>, each address might be represented as a route
so eight addresses instead of one requires the devices to support eight ti
mes more
routes, etc. If an infrastructure device's resources are exhausted, the
device might drop some IPv6 addresses from the corresponding tables,
while the address owner might still be using the address to send
traffic. This leads to traffic blackholing and a degraded customer
experience. Providing every host with one prefix allows the network to
maintain only one entry per device, while still providing the device the
ability to use an arbitrary number of addresses.
</t> </t>
<!--[rfced] We have clarified what subject "it" refers to in the sentence
below (from the Introduction) as this sentence appears far from its
original subject. Please let us know if this change alters your intended
meaning.
Original:
Second, it provides the ability to extend the network.
Current:
Second, this deployment method provides the ability to extend the network.
-->
<t>Second, this deployment model provides the ability to extend the networ
k. In IPv4, a
device that connects to the network can provide connectivity to
subtended devices by using NAT. With DHCPv6 Prefix Delegation
(DHCPv6-PD) <xref target="RFC8415"/>, such a device can similarly
extend the network, but unlike IPv4 NAT, it can provide its subtended
devices with full end-to-end connectivity.</t>
<t>Another method of deploying unique prefixes per device is documented
in <xref target="RFC8273"/>. Similarly, the standard deployment model in
cellular IPv6 networks <xref target="RFC6459"/> provides a unique prefix
to every device. However, providing a unique prefix per device is very
uncommon in enterprise-style networks, where nodes are usually connected
to broadcast segments/VLANs and each link has a single on-link prefix
assigned. This document takes a similar approach to <xref
target="RFC8273"/>, but allocates the prefix using DHCPv6-PD.</t>
<t>This document focuses on the behavior of the network. Host behavior
is not defined in this document.</t>
</section> </section>
<section> <section>
<name>Requirements Language</name> <name>Requirements Language</name>
<t>The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", <t>
"SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT The key words "<bcp14>MUST</bcp14>", "<bcp14>MUST NOT</bcp14>",
RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be "<bcp14>REQUIRED</bcp14>", "<bcp14>SHALL</bcp14>", "<bcp14>SHALL NOT</bcp14>
interpreted as described in BCP 14 <xref target="RFC2119"/> ",
<xref target="RFC8174"/> when, and only when, they appear in "<bcp14>SHOULD</bcp14>", "<bcp14>SHOULD NOT</bcp14>",
all capitals, as shown here.</t> "<bcp14>RECOMMENDED</bcp14>", "<bcp14>NOT RECOMMENDED</bcp14>",
"<bcp14>MAY</bcp14>", and "<bcp14>OPTIONAL</bcp14>" in this document are to
be
interpreted as described in BCP&nbsp;14 <xref target="RFC2119"/> <xref
target="RFC8174"/> when, and only when, they appear in all capitals, as
shown here.
</t>
</section> </section>
<section> <section>
<name>Terminology</name> <name>Terminology</name>
<t>
Node: a device that implements IPv6, <xref target="
RFC8200"/>.
</t>
<t>
Host: any node that is not a router, <xref target="
RFC8200"/>.
</t>
<t>
Client: a node which connects to a network and acquires addresses. The node may
wish to obtain addresses for its own use, or may be a router that wishes to exte
nd the network to its physical or virtual subsystems, or both. It may be either
a host or a router as defined by <xref target="RFC8200"/>.
</t>
<t> <!--[rfced] Currently, the terminology list in Section 3 is not
ND: Neighbor Discovery, <xref target="RFC4861"/>. alphabetized. May we alphabetize the list items?
</t>
<t> Original:
NUD: Neighbor Unreachability Detection, <xref target="RFC4861"/>.
</t> 3. Terminology
<t>
PIO: Prefix Information Option, <xref target="RFC4862"/>. Node: a device that implements IPv6, [RFC8200].
</t>
<t> Host: any node that is not a router, [RFC8200].
SLAAC: IPv6 Stateless Address Autoconfiguration, <xref targ
et="RFC4862"/>. Client: a node which connects to a network and acquires addresses.
</t> The node may wish to obtain addresses for its own use, or may be a
<t> router that wishes to extend the network to its physical or virtual
DHCPv6-PD: DHCPv6 (<xref target="RFC8415"/>) mechanism to delegate IPv6 prefixe subsystems, or both. It may be either a host or a router as defined
s to clients. by [RFC8200].
</t>
ND: Neighbor Discovery, [RFC4861].
NUD: Neighbor Unreachability Detection, [RFC4861].
PIO: Prefix Information Option, [RFC4862].
SLAAC: IPv6 Stateless Address Autoconfiguration, [RFC4862].
DHCPv6-PD: DHCPv6 ([RFC8415]) mechanism to delegate IPv6 prefixes to
clients.
-->
<dl>
<dt>Node:</dt><dd>a device that implements IPv6 <xref target="RFC8200"/></dd>
<dt>Host:</dt><dd>any node that is not a router <xref target="RFC8200"/></dd>
<dt>Client:</dt><dd>a node that connects to a network and acquires addresses.
The
node may wish to obtain addresses for its own use, or it may be a router that
wishes to extend the network to its physical or virtual subsystems, or
both. It may be either a host or a router as defined by <xref
target="RFC8200"/>.</dd>
<dt>ND:</dt><dd>Neighbor Discovery <xref target="RFC4861"/></dd>
<dt>NUD:</dt><dd>Neighbor Unreachability Detection <xref target="RFC4861"/></d
d>
<dt>PIO:</dt><dd>Prefix Information Option <xref target="RFC4862"/></dd>
<dt>SLAAC:</dt><dd>IPv6 Stateless Address Autoconfiguration <xref target="RFC4
862"/></dd>
<dt>DHCPv6-PD:</dt><dd>DHCPv6 Prefix Delegation <xref target="RFC8415"/>; a
mechanism to delegate IPv6 prefixes to clients.</dd>
</dl>
</section> </section>
<!--[rfced] As the text below is both the first sentence of Section 4 and the
lead-in sentence for the list that follows, would it be helpful add
additional context for the reader?
Original:
Instead of all clients on a given link forming addresses from the
same shared prefix assigned to that link:
* A device acts as DHCPv6-PD client and requests a prefix via
DHCPv6-PD by sending an IA_PD request.
Perhaps:
Instead of all clients on a given link forming addresses from the
same shared prefix assigned to that link, this deployment model follows
the design principles described below.
* A device acts as a DHCPv6-PD client and requests a prefix via
DHCPv6-PD by sending an IA_PD request.
...
-->
<section> <section>
<name>Design Principles</name> <name>Design Principles</name>
<t> <t>Instead of all clients on a given link forming addresses from the
Instead of all clients on a given link forming addresses from same shared prefix assigned to that link:</t>
the same shared prefix assigned to that link: <ul>
</t> <li>A device acts as a DHCPv6-PD client and requests a prefix via
<ul> DHCPv6-PD by sending an IA_PD request.</li>
<li> <li>The server delegates a prefix to the client and the delegated
A device acts as DHCPv6-PD client and requests a pref prefix is installed into the routing table of the first-hop router as
ix via DHCPv6-PD by sending an IA_PD request. a route pointing to the client's link-local address. The first-hop
</li> router can act as a DHCPv6 relay and snoop DHCPv6 Reply messages from
<li> an off-link DHCPv6 server, or it can act as a DHCPv6 server itself. In
The server delegates a prefix to the client and the d both cases, it can install the route locally, and if the network is
elegated prefix is installed into the routing table of the first-hop router as a running a dynamic routing protocol, distribute the route or the entire
route pointing to the client's link-local address. The first-hop router can act prefix pool into the protocol.</li>
as a DHCPv6 relay and snoop DHCPv6 Reply messages from an off-link DHCPv6 serve <li>For the router and all other infrastructure devices, the delegated
r, or it can act as a DHCPv6 server itself. In both cases it can install the rou prefix is considered off-link, so traffic to that prefix does not
te locally, and if the network is running a dynamic routing protocol, distribute trigger any ND packets, other than the minimum ND required to sustain
the route or the entire prefix pool into the protocol.</li> Neighbor Unreachability Detection (NUD) for the client's link-local
<li>For the router and all other infrastructure devices, address.</li>
the delegated prefix is considered off-link, so traffic to that prefix does not <li>The device can use the delegated prefix in various ways. For
trigger any ND packets, other than the minimum ND required to sustain Neighbor U example, it can form addresses, as described in requirement WAA-7 of
nreachability Detection (NUD) for the client's link-local address. <xref target="RFC7084"/>. It can also extend the network, as described
</li> in <xref target="RFC7084"/> or <xref target="RFC7278"/>.</li>
<li> </ul>
The device can use the delegated prefix in various ways. For exa
mple, it can form addresses, as described in <xref target="RFC7084"/> requiremen <t>An example scenario is shown in <xref target="fig1"/>. Note that the pref
t WAA-7. It can also extend the network, as described in <xref target="RFC7084"/ ix lengths
> or <xref target="RFC7278"/>. used in the example are /64 because that is the prefix length currently
</li> supported by SLAAC and is not otherwise required by the proposed
</ul> deployment model.</t>
<t>
An example scenario is shown in Figure 1. Note that the prefix lengths used in t
he example are /64 because that is the prefix length currently supported by SLAA
C and is not otherwise required by the proposed deployment model.
</t>
<figure anchor="fig1"> <figure anchor="fig1">
<name> <name>An Example Topology with Two First-Hop Routers</name>
An Example Topology with Two First-Hop Routers.
</name> <!--[rfced] Please review the following questions about the ASCII and SVG
artwork in Figure 1:
a. There are some inconsistencies between the SVG and ASCII artwork (see a few
below). Please review and let us know how to update so these artworks match.
i. Tethered device (ASCII) v. Tethered device1 and Tethered device2 (SVG)
ii. Text appears to be cut off in the "virtual system" boxes and the "Legacy
client" box (ASCII)
iii. Router Advertisement containing PIO (ASCII) v. Router Advertisement PIO
(SVG)
iv. Legacy (no DHCPv6-PD) client B (ASCII) v. legacy client B no DHCPv6-PD
support (SVG)
b. The ASCII art is too wide; it extends beyond the 72 character limit. How may
we adjust?
-->
<artset> <artset>
<artwork type="svg"> <artwork type="svg">
<svg xmlns="http://www.w3.org/2000/svg" version="1.1" height="928" width="856" v iewBox="0 0 856 928" class="diagram" text-anchor="middle" font-family="monospace " font-size="13px"> <svg xmlns="http://www.w3.org/2000/svg" version="1.1" height="928" width="856" v iewBox="0 0 856 928" class="diagram" text-anchor="middle" font-family="monospace " font-size="13px">
<path d="M 8,128 L 8,288" fill="none" stroke="black"/> <path d="M 8,128 L 8,288" fill="none" stroke="black"/>
<path d="M 8,736 L 8,912" fill="none" stroke="black"/> <path d="M 8,736 L 8,912" fill="none" stroke="black"/>
<path d="M 24,336 L 24,432" fill="none" stroke="black"/> <path d="M 24,336 L 24,432" fill="none" stroke="black"/>
<path d="M 24,816 L 24,880" fill="none" stroke="black"/> <path d="M 24,816 L 24,880" fill="none" stroke="black"/>
<path d="M 48,48 L 48,176" fill="none" stroke="black"/> <path d="M 48,48 L 48,176" fill="none" stroke="black"/>
<path d="M 48,224 L 48,336" fill="none" stroke="black"/> <path d="M 48,224 L 48,336" fill="none" stroke="black"/>
skipping to change at line 322 skipping to change at line 446
<text x="116" y="852">2001:db8:ddd0:1::f00</text> <text x="116" y="852">2001:db8:ddd0:1::f00</text>
<text x="320" y="852">2001:db8:ddd0:1::2345</text> <text x="320" y="852">2001:db8:ddd0:1::2345</text>
<text x="120" y="868">2001:db8:ddd0:1::cafe</text> <text x="120" y="868">2001:db8:ddd0:1::cafe</text>
<text x="316" y="868">2001:db8:ddd0:1::abc</text> <text x="316" y="868">2001:db8:ddd0:1::abc</text>
<text x="572" y="884">Tethered device1</text> <text x="572" y="884">Tethered device1</text>
<text x="756" y="884">Tethered device2</text> <text x="756" y="884">Tethered device2</text>
<text x="568" y="900">2001:db8:ddd0:2::5555</text> <text x="568" y="900">2001:db8:ddd0:2::5555</text>
<text x="764" y="900">2001:db8:ddd0:2::6666</text> <text x="764" y="900">2001:db8:ddd0:2::6666</text>
</g> </g>
</svg> </svg>
</artwork> </artwork>
<artwork type="ascii-art">
<artwork type="ascii-art">
<![CDATA[ <![CDATA[
+-----------------------------------------------------------------------------+ +-----------------------------------------------------------------------------+
| DHCPv6 Servers | | DHCPv6 Servers |
| Pool 2001:db8:ddd0::/48 for clients on 2001:db8:c001::/64 link | | Pool 2001:db8:ddd0::/48 for clients on 2001:db8:c001::/64 link |
+---------------+-------------------------+-----------------------------+-----+ +---------------+-------------------------+-----------------------------+-----+
^ | | ^ | ^ | | ^ |
| | | | | | | | | |
+-------+------+-------------------------+----------------------+------+-----+ +-------+------+-------------------------+----------------------+------+-----+
|DHCPv6 | | IPv6 Network DHCPv6 | | | |DHCPv6 | | IPv6 Network DHCPv6 | | |
|Relay-Forward | Relay-Forward | | |Relay-Forward | Relay-Forward | |
skipping to change at line 380 skipping to change at line 504
| | 2001:db8:ddd0:1::caf| |2001:db8:ddd0:1::a| | | Tethered device | | | 2001:db8:ddd0:1::caf| |2001:db8:ddd0:1::a| | | Tethered device |
| +---------------------+ +------------------+ | |2001:db8:ddd0:2::6666| | +---------------------+ +------------------+ | |2001:db8:ddd0:2::6666|
| | +---------------------+ | | +---------------------+
+-----------------------------------------------+ +-----------------------------------------------+
]]> ]]>
</artwork> </artwork>
</artset> </artset>
</figure> </figure>
</section> </section>
<section> <section>
<name>Applicability and Limitations</name> <name>Applicability and Limitations</name>
<t>
Delegating a unique prefix per client provides all the benefi <t>Delegating a unique prefix per client provides all the benefits of both
ts of both SLAAC and DHCPv6 address allocation, but at the cost of greater addre SLAAC and DHCPv6 address allocation, but at the cost of greater address-spac
ss space usage. e usage. This design would substantially benefit some networks (see
This design would substantially benefit some networks (see <x <xref target="benefits"/>) in which the additional cost of an additional
ref target="benefits"/>), in which the additional cost of an additional service service (such as DHCPv6 Prefix Delegation) and allocation of a larger amount
(DHCPv6 prefix delegation) and allocating a larger amount of address space can e of
asily be justified. address space can easily be justified. Examples of such networks include
Examples of such networks include but are not limited to: but are not limited to:</t>
</t>
<ul> <ul>
<li> <li>Large-scale networks where even three to five addresses per cli
Large-scale networks where even 3-5 addresses per cli ent
ent might introduce scalability issues. might introduce scalability issues.</li>
</li> <li>Networks with a high number of virtual hosts, so physical
<li> devices require multiple addresses.</li>
Networks with a high number of virtual hosts, so phys <li>Managed networks where extensive troubleshooting, device
ical devices require multiple addresses. traffic logging, or forensics might be required.</li>
</li>
<li>
Managed networks where extensive troubleshooting, dev
ice traffic logging, or forensics might be required.
</li>
</ul> </ul>
<t> <t>In smaller networks, such as home networks or small
In smaller networks, such as home networks or small enterpris enterprises with smaller address space and a lower number of
es, with smaller address space and lower number of clients, SLAAC is a simpler a clients, SLAAC is a simpler and often preferred option.</t>
nd often preferred option.
</t>
</section> </section>
<section> <section>
<name>Routing and Addressing Considerations</name> <name>Routing and Addressing Considerations</name>
<section> <section>
<name>Prefix Pool Allocation</name> <name>Prefix Pool Allocation</name>
<t>One simple deployment model is to assign a dedicated prefix pool <t>One simple deployment model is to assign a dedicated prefix pool to
to each link. The prefixes from each link's pool are only issued to requesting c each link. The prefixes from each link's pool are only issued to
lients on the link, and if clients move to another link they will obtain a prefi requesting clients on the link; if clients move to another link,
x from the pool associated with the new link (see <xref target="mobility"/>).</t they will obtain a prefix from the pool associated with the new link
> (see <xref target="mobility"/>).</t>
<t>This is very similar to how address pools are allocated when usin
g DHCP to assign individual addresses (e.g., DHCPv4 or DHCPv6 IA_NA), where each
link has a dedicated pool of addresses, and clients on the link obtain addresse
s from the pool. In this model, the network can route the entire pool to the lin
k's first-hop routers, and the routers do not need to advertise individual deleg
ated prefixes into the network's dynamic routing protocol.</t>
<t>Other deployment models, such as prefix pools shared over multipl <t>This is very similar to how address pools are allocated when using
e links or routers, are possible, but not described in this document.</t> DHCP to assign individual addresses (e.g., DHCPv4 or DHCPv6 IA_NA),
</section> where each link has a dedicated pool of addresses, and clients on the
link obtain addresses from the pool. In this model, the network can
route the entire pool to the link's first-hop routers, and the routers
do not need to advertise individual delegated prefixes into the
network's dynamic routing protocol.</t>
<t>Other deployment models, such as prefix pools shared over multiple
links or routers, are possible but are not described in this
document.</t>
</section>
<section> <section>
<name> <name>First-Hop Router Requirements</name>
First-Hop Router Requirements <t>In large networks, DHCPv6 servers are usually centralized and reached
</name> via DHCPv6 relays co-located with the first-hop routers. To delegate IPv6
<t> prefixes to clients, the first hop routers need to implement DHCPv6 relay
In large networks, DHCPv6 servers are usually centralized, an functions and meet the requirements defined in <xref target="RFC8987"/>.
d reached via DHCPv6 relays co-located with the first-hop routers. In particular, per <xref target="RFC8987" sectionFormat="of" section="4.2"/>
To delegate IPv6 prefixes to clients, the first hop routers n , the first-hop
eed to implement DHCPv6 relay functions and meet the requirements defined in <xr router must maintain a local routing table that contains all prefixes
ef target="RFC8987"/>. delegated to clients.</t>
In particular, per <xref target="RFC8987" section="4.2"/>, th
e first-hop router must maintain a local routing table that contains all prefixe
s delegated to clients.</t>
<t>With the first-hop routers performing DHCPv6 relay functions, <t>With the first-hop routers performing DHCPv6 relay functions, the
the proposed design neither requires any subsequent relays in the path nor intro proposed design neither requires any subsequent relays in the path nor
duces any requirements (e.g., snooping) to such subsequent relays, if they are d introduces any requirements (e.g., snooping) for such subsequent relays, if
eployed. </t> they are deployed.</t>
<t> <t>To ensure that routes to the delegated prefixes are preserved even if a
To ensure that routes to the delegated prefixes are preserved even if a relay is relay is rebooted or replaced, the operator <bcp14>MUST</bcp14> ensure
rebooted or replaced, the operator MUST ensure that all relays in the network i that all relays in the network infrastructure support DHCPv6 Bulk
nfrastructure support DHCPv6 Bulk Leasequery as defined in <xref target="RFC5460 Leasequery as defined in <xref target="RFC5460"/>. While <xref
"/>. target="RFC8987" section="4.3" sectionFormat="of"/> lists keeping active
prefix delegations in persistent storage as an alternative to DHCPv6 Bulk
Leasequery, relying on persistent storage has the following drawbacks:
</t>
While Section 4.3 of <xref target="RFC8987"/> lists keeping active prefix delega <!--[rfced] For ease of the reader, how may we add additional context and/or
tions in persistent storage as an alternative to DHCPv6 Bulk Leasequery, relying verbs to clarify the items in parentheses below?
on persistent storage has the following drawbacks:
</t> Original:
<ul>
<li> * In a network with multiple relays, network state can change
In a network with multiple relays, network state can change significantly while significantly while the relay is rebooting (new prefixes
the relay is rebooting (new prefixes delegated, some prefixes expiring etc). delegated, some prefixes expiring etc).
</li>
<li> Perhaps:
Persistent storage might not be preserved if the router is physically replaced.
</li> * In a network with multiple relays, network state can change
</ul> significantly while the relay is rebooting (new prefixes might
<t>Another mechanism for first-hop routers to obtain information be delegated or some prefixes might be expiring, etc).
about delegated prefixes is by using Active Leasequery <xref target="RFC7653"/>,
though this is not yet widely supported.</t>
-->
<ul>
<li>In a network with multiple relays, network state can change
significantly while the relay is rebooting (new prefixes delegated, some
prefixes expiring, etc.).</li>
<li>Persistent storage might not be preserved if the router is
physically replaced.</li>
</ul>
<t>Another mechanism for first-hop routers to obtain information about
delegated prefixes is by using Active Leasequery <xref target="RFC7653"/>,
though this is not yet widely supported.</t>
</section> </section>
<section anchor="mult_relays"> <section anchor="mult_relays">
<name> <name>Topologies with Multiple First-Hop Routers</name>
Topologies with Multiple First-Hop Routers
</name> <t>In a topology with redundant first-hop routers, all the routers need to
<t> relay DHCPv6 traffic, install the delegated prefixes into their routing
In a topology with redundant first-hop routers, all the route tables and, if needed, advertise those prefixes to the network.</t>
rs need to relay DHCPv6 traffic, install the delegated prefixes into their routi
ng tables and, if needed, advertise those prefixes to the network.</t> <t>If the first-hop routers obtain information about delegated prefixes by
<t>If the first-hop routers obtain information about delegated pr snooping DHCPv6 Reply messages sent by the server, then all the first-hop
efixes by snooping DHCPv6 Reply messages sent by the server, then all the first- routers must be able to snoop these messages. This is possible if the
hop routers must be able to snoop these messages. This is possible if the client client multicasts the DHCPv6 messages it sends to the server. The server
multicasts the DHCPv6 messages it sends to the server. The server will receive will receive one copy of the client message through each first-hop relay,
one copy of the client message through each first-hop relay, and will reply unic and will reply unicast to each of them via the relay (or chain of relays)
ast to each of them via the relay (or chain of relays) from which it received th from which it received the message. Thus, all first-hop relays will be
e message. Thus, all first-hop relays will be able to snoop the replies. Per <xr able to snoop the replies. Per <xref target="RFC8415" sectionFormat="of" sec
ef target="RFC8415" section="14"/>, clients always use multicast unless the serv tion="14"/>,
er uses the Server Unicast option to explicitly allow unicast communication (<xr clients always use multicast unless the server uses the Server Unicast
ef target="RFC8415" section="21.12" sectionFormat="comma"/>). Therefore, in topo option to explicitly allow unicast communication (<xref target="RFC8415"
logies with multiple first-hop routers, the DHCPv6 servers MUST be configured no section="21.12" sectionFormat="comma"/>). Therefore, in topologies with
t to use the Server Unicast option. It should be noted that <xref target="I-D.ie multiple first-hop routers, the DHCPv6 servers <bcp14>MUST</bcp14> be
tf-dhc-rfc8415bis"/> deprecates the Server Unicast option precisely because it i configured not to use the Server Unicast option. It should be noted that
s not compatible with topologies with multiple first-hop relays. <xref target="I-D.ietf-dhc-rfc8415bis"/> deprecates the Server Unicast
</t> option precisely because it is not compatible with topologies with
<t>To recover from crashes or reboots, relays can use Bulk Leasequery multiple first-hop relays.</t>
or Active Leasequery to issue a QUERY_BY_RELAY_ID with the ID(s) of the other r
elay(s), as configured by the operator. Additionally, some vendors provide vendo <t>To recover from crashes or reboots, relays can use Bulk Leasequery or
r-specific mechanisms to synchronize state between DHCP relays.</t> Active Leasequery to issue a QUERY_BY_RELAY_ID with the ID(s) of the other
relay(s), as configured by the operator. Additionally, some vendors
provide vendor-specific mechanisms to synchronize state between DHCP
relays.</t>
</section> </section>
<section> <section>
<name>On-link Communication </name> <name>On-Link Communication </name>
<t> <t>For security reasons, some networks block on-link device-to-device
For security reasons, some networks block on-link device-to-device traffic at la traffic at Layer 2 to prevent communication between clients on the same
yer 2 to prevent communication between clients on the same link. link. In this case, delegating a prefix to each client doesn't affect
In this case, delegating a prefix to each client doesn't affect traffic flows, a traffic flows, as all traffic is sent to the first-hop router anyway.
s all traffic is sent to the first-hop router anyway. Depending on the network security policy, the router may allow or drop
Depending on the network security policy, the router may allow or drop the traff the traffic.</t>
ic.
</t>
<t>
If the network does allow peer-to-peer communication, the PIO for the on-link pr
efix is usually advertised with the L-bit set to 1 <xref target="RFC4861"/>.
As a result, all addresses from that prefix are considered on-link, and traffic
to those destinations is sent directly (not via routers).
If such a network delegates prefixes to clients (as described in this document),
then each client will consider other client's destination addresses to be off-l
ink, because those addresses are from the delegated prefixes and are no longer w
ithin the on-link prefix.
When a client sends traffic to another client, packets will initially be sent to
the default router.
The router will respond with an ICMPv6 redirect message (Section 4.5 of <xref ta
rget="RFC4861" />). If the client receives and accepts the redirect, then traffi
c can flow directly from device to device.
Therefore the administrator deploying the solution described in this document SH
OULD ensure that the first-hop routers can send ICMPv6 redirects (the routers ar
e configured to do so and the security policies permit those messages).
</t>
<t>If the network does allow peer-to-peer communication, the PIO for the
on-link prefix is usually advertised with the L-bit set to 1 <xref
target="RFC4861"/>. As a result, all addresses from that prefix are
considered on-link, and traffic to those destinations is sent directly
(not via routers). If such a network delegates prefixes to clients (as
described in this document), then each client will consider other
client's destination addresses to be off-link, because those addresses
are from the delegated prefixes and are no longer within the on-link
prefix. When a client sends traffic to another client, packets will
initially be sent to the default router. The router will respond with
an ICMPv6 redirect message (<xref target="RFC4861" section="4.5"
sectionFormat="of"/>). If the client receives and accepts the redirect,
then traffic can flow directly from device to device. Therefore, the
administrator deploying the solution described in this document
<bcp14>SHOULD</bcp14> ensure that the first-hop routers can send ICMPv6
redirects (the routers are configured to do so and the security policies
permit those messages).</t>
</section> </section>
</section> </section>
<section> <section>
<name>DHCPv6-PD Server Considerations</name> <name>DHCPv6-PD Server Considerations</name>
<t>
This document does not introduce any changes to the DHCPv6 protocol itself.
However, for the proposed solution to work correctly, the DHCPv6-PD server needs
to be configured as follows:
</t>
<ul>
<li>
The server MUST follow <xref target="RFC8168"/> recom
mendations on processing prefix-length hints.
</li>
<li>
The server MUST provide a prefix short enough for the client to extend the netwo
rk to at least one interface, and allow nodes on that interface to obtain addres
ses via SLAAC.
The server MAY provide more address space to clients that ask for it, either by
delegating multiple such prefixes, or by delegating a single prefix of a shorter
length. It should be noted that <xref target="RFC8168"/> allows the server to p
rovide a prefix shorter than the prefix-length hint value received from the clie
nt.
</li>
<li>
If the server receives the same SOLICIT message from
the same client multiple times through multiple relays, it MUST reply to all of
them with the same prefix(es).
This ensures that all the relays will correctly confi
gure routes to the delegated prefixes.
</li>
<li>
The server MUST NOT send the Server Unicast option to
the client unless the network topology guarantees that no client is connected t
o a link with multiple relays (see <xref target="mult_relays"/>).
</li>
<li>
In order to ensure uninterrupted connectivity when a
first-hop router crashes or reboots, the server MUST support Bulk Leasequery or
Active Leasequery.
</li>
</ul>
<t>
As most operators have some experience with IPv4, they can use a similar approac
h for choosing the pool size and the timers (such as T1/T2 timers).
In particular the following factors shall be taken into account:
</t>
<ul>
<li>
the expected maximum number of clients;
</li>
<li>
average duration of a client connection;
</li>
<li>
how mobile the clients are (a network where all clients are connected to a sing
le wired VLAN might choose longer timers than a network where clients can switch
between multiple wireless SSIDs);
</li>
<li>
expected level of recurring clients (for example, a corporate authenticated Wi-F
i network might be using longer timers than an open public Wi-Fi).
</li>
</ul>
<t> <t>This document does not introduce any changes to the DHCPv6 protocol
DHCPv6 servers that delegate prefixes can interface with Dynamic DNS infrastruct itself. However, for the proposed solution to work correctly, the
ure to automatically populate reverse DNS, DHCPv6-PD server needs to be configured as follows:</t>
similarly to what is described in section 2.5.2 of RFC <xref target="RFC8501"/>. <ul>
Networks that also wish to populate forward DNS cannot do so <li>The server <bcp14>MUST</bcp14> follow recommendations from <xref
automatically based only on DHCPv6 prefix delegation transactions, but they can target="RFC8168"/> on processing prefix-length hints.</li>
do so in other ways, such as by supporting <li>The server <bcp14>MUST</bcp14> provide a prefix short enough for the
DHCPv6 address registration as described in <xref target="I-D.ietf-dhc-addr-noti client to extend the network to at least one interface and allow nodes
fication"/>. on that interface to obtain addresses via SLAAC. The server
</t> <bcp14>MAY</bcp14> provide more address space to clients that ask for
it, either by delegating multiple such prefixes, or by delegating a
single prefix of a shorter length. It should be noted that <xref
target="RFC8168"/> allows the server to provide a prefix shorter than
the prefix-length hint value received from the client.</li>
<li>If the server receives the same Solicit message from the same
client multiple times through multiple relays, it <bcp14>MUST</bcp14>
reply to all of them with the same prefix(es). This ensures that all
the relays will correctly configure routes to the delegated prefixes.</li>
<li>The server <bcp14>MUST NOT</bcp14> send the Server Unicast option to
the client unless the network topology guarantees that no client is
connected to a link with multiple relays (see <xref
target="mult_relays"/>).</li>
<li>In order to ensure uninterrupted connectivity when a first-hop
router crashes or reboots, the server <bcp14>MUST</bcp14> support Bulk
Leasequery or Active Leasequery.</li>
</ul>
<t>Some additional recommendations driven by security and privacy considerations <t>As most operators have some experience with IPv4, they can use a
are discussed in <xref target="Security"/> and <xref target="privacy"/>.</t> similar approach for choosing the pool size and the timers (such as T1/T2
timers). In particular, the following factors shall be taken into account:<
/t>
<!--[rfced] FYI - We have slightly adjusted the list items below for
consistency. Please review and let us know if these changes alter your
intended meaning.
Original:
In particular the following factors shall be taken
into account:
* the expected maximum number of clients;
* average duration of a client connection;
* how mobile the clients are (a network where all clients are
connected to a single wired VLAN might choose longer timers than a
network where clients can switch between multiple wireless SSIDs);
* expected level of recurring clients (for example, a corporate
authenticated Wi-Fi network might be using longer timers than an
open public Wi-Fi).
Current:
In particular, the following factors shall be taken
into account:
* the expected maximum number of clients;
* the average duration of a client connection;
* the mobility of the clients (for example, a network where all
clients are connected to a single wired VLAN might choose longer
timers than a network where clients can switch between multiple
wireless SSIDs);
* the expected level of recurring clients (for example, a corporate
authenticated Wi-Fi network might be using longer timers than an
open public Wi-Fi).
-->
<ul>
<li>the expected maximum number of clients;</li>
<li>the average duration of a client connection;</li>
<li>the mobility of the clients (for example, a network where all clients
are
connected to a single wired VLAN might choose longer timers than a
network where clients can switch between multiple wireless SSIDs);</li>
<li>the expected level of recurring clients (for example, a corporate
authenticated Wi-Fi network might be using longer timers than an open
public Wi-Fi).</li>
</ul>
<!-- [rfced] RFC 8501 does not have a section 2.5.2. Perhaps the text should re
fer to section 2.3.2 (see https://www.rfc-editor.org/rfc/rfc8501#section-2.3.2)?
Original:
DHCPv6 servers that delegate prefixes can interface with Dynamic DNS
infrastructure to automatically populate reverse DNS, similarly to
what is described in section 2.5.2 of RFC [RFC8501].
-->
<t>DHCPv6 servers that delegate prefixes can interface with Dynamic DNS
infrastructure to automatically populate reverse DNS, similarly to what is
described in <xref target="RFC8501" sectionFormat="of"
section="2.5.2"/>. Networks that also wish to populate forward DNS cannot
do so automatically based only on DHCPv6 prefix delegation transactions,
but they can do so in other ways, such as by supporting DHCPv6 address
registration as described in <xref
target="I-D.ietf-dhc-addr-notification"/>.</t>
<t>Some additional recommendations driven by security and privacy
considerations are discussed in <xref target="Security"/> and <xref
target="privacy"/>.</t>
</section> </section>
<section> <section>
<name>Prefix Length Considerations</name> <name>Prefix Length Considerations</name>
<t>Delegating a prefix of sufficient size to use SLAAC allows the cli
ent to extend the network, providing limitless addresses to IPv6 nodes connected
to it (e.g., virtual machines, tethered devices), because all IPv6 hosts are re
quired to support SLAAC <xref target="RFC8504"/>. Additionally, even clients tha
t support other forms of address assignment require SLAAC for some functions, su
ch as forming dedicated addresses for the use of 464XLAT (see Section 6.3 of <xr
ef target="RFC6877"/>).</t>
<t>At the time of writing the only prefix size that will allow device <t>Delegating a prefix of sufficient size to use SLAAC allows the client
s to use SLAAC is 64 bits. Also, as noted in <xref target="RFC7421"/>, using an to extend the network, providing limitless addresses to IPv6 nodes
IID shorter than 64 bits and a subnet prefix longer than 64 bits is outside the connected to it (e.g., virtual machines or tethered devices), because all
current IPv6 specifications. IPv6 hosts are required to support SLAAC <xref
Choosing longer prefixes would require the client and any connected system to us target="RFC8504"/>. Additionally, even clients that support other forms of
e other address assignment mechanisms. address assignment require SLAAC for some functions, such as forming
This would limit the applicability of the proposed solution, as other mechanisms dedicated addresses for the use of 464XLAT (see <xref
are not currently supported by many hosts. target="RFC6877" section="6.3" sectionFormat="of"/>).</t>
</t>
<t>For the same reasons, a prefix length of /64 or shorter is require <t>At the time of writing, the only prefix size that will allow devices to
d to extend the network using <xref target="RFC7084"/> (see requirement L-2), an use SLAAC is 64 bits. Also, as noted in <xref target="RFC7421"/>, using an in
d a prefix length of /64 is required to provide global connectivity for stub net terface identifier (IID) shorter than 64 bits and a subnet prefix longer than 64
works as per <xref target="I-D.ietf-snac-simple"/>.</t> bits is outside
the current IPv6 specifications. Choosing longer prefixes would require
the client and any connected system to use other address assignment
mechanisms. This would limit the applicability of the proposed solution,
as other mechanisms are not currently supported by many hosts.</t>
<t> <t>For the same reasons, a prefix length of /64 or shorter is required to
Assigning a prefix of sufficient size to support SLAAC is possible o extend the network as described in <xref target="RFC7084"/> (see requirement
n large networks. In general, any network that numbers clients from an IPv4 pref L-2),
ix of length X (e.g., X=/18, X=/24), would require an IPv6 prefix of length X+32 and a prefix length of /64 is required to provide global connectivity for
(e.g., X=/40, X=/56) to provide a /64 prefix to every device. stub networks as per <xref target="I-D.ietf-snac-simple"/>.</t>
As an example, <xref target="RFC7934" section="9.2"/> suggests that e
ven a very large network that assigns every single one of the 16 million IPv4 ad
dresses in 10.0.0.0/8 would only need an IPv6 /40. A /40 prefix is a small amoun
t of address space: there are 32 times more /40s in the current IPv6 unicast ran
ge 2000::/3 than there are IPv4 addresses.
Existing sites that currently use a /48 prefix cannot support more th
an 64k clients in this model without renumbering, though many networks of such s
ize have LIR status and can justify bigger address blocks.
</t>
<t>Note that assigning a prefix of sufficient size to support SLAAC does <t>Assigning a prefix of sufficient size to support SLAAC is possible on
not require that subtended nodes use SLAAC; they can use other address assignme large networks. In general, any network that numbers clients from an IPv4
nt mechanisms as well.</t> prefix of length X (e.g., X=/18, X=/24) would require an IPv6 prefix of
</section> length X+32 (e.g., X=/40, X=/56) to provide a /64 prefix to every device.
As an example, <xref target="RFC7934" section="9.2"/> suggests that even a
very large network that assigns every single one of the 16 million IPv4
addresses in 10.0.0.0/8 would only need an IPv6 /40. A /40 prefix is a
small amount of address space: there are 32 times more /40s in the current
IPv6 unicast range 2000::/3 than there are IPv4 addresses. Existing sites
that currently use a /48 prefix cannot support more than 64k clients in
this model without renumbering, though many networks of such size have Local
Internet Registry (LIR) status and can justify bigger address blocks.</t>
<t>Note that assigning a prefix of sufficient size to support SLAAC does
not require that subtended nodes use SLAAC; they can use other address
assignment mechanisms as well.</t>
</section>
<section anchor="mobility"> <section anchor="mobility">
<name>Client Mobility</name> <name>Client Mobility</name>
<t>
As per Section 18.2.12 of <xref target="RFC8415"/>, when the client moves to a n
ew link, it MUST initiate a Rebind/Reply message exchange. Therefore when the cl
ient moves between network attachment points it would refresh its delegated pref
ix the same way it refreshes addresses assigned (via SLAAC or DHCPv6 IA_NA) from
a shared on-link prefix:
</t>
<ul>
<li> <t>As per <xref target="RFC8415" section="18.2.12" sectionFormat="of"/>, when
When a client moves from between different attachment points on the same link (e the client moves to a new link, it <bcp14>MUST</bcp14> initiate a Rebind/Reply
.g., roams between two APs while connected to the same SSID or moves between two message exchange. Therefore, when the client moves between network attachment
switchports belonging to the same VLAN), the delegated prefix does not change, points, it would refresh its delegated prefix the same way it refreshes
and the first-hop routers have a route for the prefix with the nexthop set to th addresses assigned (via SLAAC or DHCPv6 IA_NA) from a shared on-link prefix:</t>
e client link-local address on that link. As per requirement S-2 (Section 4.3 of
<xref target="RFC8987"/>), the DHCPv6-relays (the first-hop routers) MUST retai
n the route for the delegating prefix until the route is released or removed due
to expiring DHCP timers. Therefore, if the client reconnects to the same link,
the prefix doesn't change.
</li>
<li> <ul>
When a client moves to a different link, the DHCPv6 server provides the client w <li>When a client moves between different attachment points on the same
ith a new prefix, so the behaviour is consistent with SLAAC or DHCPv6-assigned a link (e.g., roams between two APs while connected to the same SSID or moves
ddresses, which are also different on the new link. between two switch ports belonging to the same VLAN), the delegated prefix
</li> does not change, and the first-hop routers have a route for the prefix with
the nexthop set to the client link-local address on that link. As per
requirement S-2 in <xref target="RFC8987" section="4.3" sectionFormat="of"/>,
the DHCPv6-relays
(the first-hop routers) <bcp14>MUST</bcp14> retain the route for the
delegating prefix until the route is released or removed due to expiring
DHCP timers. Therefore, if the client reconnects to the same link, the
prefix doesn't change.</li>
<li>When a client moves to a different link, the DHCPv6 server provides the
client with a new prefix, so the behavior is consistent with SLAAC or
DHCPv6-assigned addresses, which are also different on the new link.</li>
</ul> </ul>
<t> <t>In theory, DHCPv6 servers can delegate the same prefix to the same client
In theory, DHCPv6 servers can delegate the same prefix to the same client even i even if the client changes the attachment points. However, while allowing the
f the client changes the attachment points. client to keep the same prefix while roaming between links might provide some
However, while allowing the client to keep the same prefix while roaming between benefits for the client, it is not feasible without changing DHCPv6 relay
links might provide some benefits for the client, it is not feasible without ch behavior: after the client moves to a new link, the DHCPv6 relays would
anging DHCPv6 relay behaviour: after the client moves to a new link, the DHCPv6 retain the route pointing to the client's link-local address on the old link
relays would retain the route pointing to the client's link-local address on the for the duration of DHCPv6 timers (see requirement S-2, <xref target="RFC8987"
old link for the duration of DHCPv6 timers (see requirement S-2, Section 4.3 of section="4.3" sectionFormat="of"/>). As a result, the first-hop routers would
<xref target="RFC8987"/>). have two routes for the same prefix pointing to different links, causing
As a result, the first-hop routers would have two routes for the same prefix poi connectivity issues for the client.</t>
nting to different links, causing connectivity issues for the client.
</t> <t>It should be noted that addressing clients from a shared on-link prefix
<t>It should be noted that addressing clients from a shared on-link prefix also also does not allow clients to keep addresses while roaming between links, so
does not allow clients to keep addresses while roaming between links, so the pro the proposed solution is not different in that regard. In addition to that,
posed solution is not different in that regard. In addition to that, quite often different links often have different security policies applied (for
different links have different security policies applied (for example, corporat example, corporate internal networks versus guest networks), hence clients on
e internal network vs guest network), hence clients on different links need to u different links need to use different prefixes.</t>
se different prefixes.
</t>
</section> </section>
<section anchor="savi"> <!--[rfced] We were unable to find the phrase "unicast Reverse Path
<name>Antispoofing and SAVI Interaction</name> Forwarding (uRPF)" in RFC 3704. However, other types of Reverse Path
<t> Forwarding do appear (e.g., Loose RPF, Feasible RPF, Strict RPF). Are any
Enabling the unicast Reverse Path Forwarding (uRPF, <xref tar updates needed to the text or citation below?
get="RFC3704"/>) on the first-hop router interfaces towards clients provides the
first layer of defense against spoofing. Original:
A spoofed packet sent by a malicious client would be dropped
by the router unless the spoofed address belongs to a prefix delegated to anothe Enabling the unicast Reverse Path Forwarding (uRPF, [RFC3704]) on the
r client on the same interface. first-hop router interfaces towards clients provides the first layer
Therefore the malicious client can only spoof addresses alrea of defense against spoofing.
dy delegated to another client on the same link or another client link-local add
ress. -->
</t>
<t> <section anchor="savi">
Source Address Validation Improvement (SAVI, <xref target="RF <name>Antispoofing and SAVI Interaction</name>
C7039"/>) provides more reliable protection against address spoofing. <!-- [rfced] Please clarify "another client link-local address". Should this be
Administrators deploying the proposed solution on SAVI-enable "another client's link-local address"?
d infrastructure SHOULD ensure that SAVI perimeter devices support DHCPv6-PD sno
oping to create the correct binding for the delegated prefixes (see <xref target Original:
="RFC7513"/>). Therefore the malicious client can only spoof addresses
Using FCFS SAVI (<xref target="RFC6620"/>) for protecting lin already delegated to another client on the same link or another
k-local addresses and creating SAVI bindings for DHCPv6-PD assigned prefixes wou client link-local address.
ld prevent spoofing. -->
</t>
<t>Enabling unicast Reverse Path Forwarding (uRPF) <xref
target="RFC3704"/> on the first-hop router interfaces towards clients
provides the first layer of defense against spoofing. A spoofed packet
sent by a malicious client would be dropped by the router unless the
spoofed address belongs to a prefix delegated to another client on the
same interface. Therefore, the malicious client can only spoof addresses
already delegated to another client on the same link or another client
link-local address.</t>
<t>Source Address Validation Improvement (SAVI) <xref target="RFC7039"/>
provides more reliable protection against address spoofing.
Administrators deploying the proposed solution on SAVI-enabled
infrastructure <bcp14>SHOULD</bcp14> ensure that SAVI perimeter devices
support DHCPv6-PD snooping to create the correct binding for the delegated
prefixes (see <xref target="RFC7513"/>). Using FCFS SAVI <xref
target="RFC6620"/> to protect link-local addresses and create SAVI
bindings for DHCPv6-PD assigned prefixes would prevent spoofing.</t>
<t>Some infrastructure devices do not implement SAVI as defined in <xref
target="RFC7039"/>; instead, they perform other forms of address tracking an
d
snooping for security or performance improvement purposes (e.g., ND
proxy). This is very common behavior for wireless devices (such as access p
oints
and controllers). Administrators <bcp14>SHOULD</bcp14> ensure that such
devices are able to snoop DHCPv6-PD packets so the traffic from the
delegated prefixes is not dropped.</t>
<t>It should be noted that using DHCPv6-PD makes it harder for an attacker
to select the spoofed source address. When all clients are using the same
shared link to form addresses, the attacker might learn addresses used by
other clients by listening to multicast Neighbor Solicitations and
Neighbor Advertisements. In DHCPv6-PD environments, however, the
attacker can only learn about other clients' global addresses by
listening to multicast DHCPv6 messages, which are not transmitted so
often, and may not be received by the client at all because they are sent
to multicast groups that are specific to DHCPv6 servers and relays.</t>
<t>
Some infrastructure devices do not implement SAVI as defined in <xref target="RF
C7039"/> but perform other forms of address tracking and snooping for security o
r performance improvement purposes (e.g., ND proxy).
This is very common behaviour for wireless devices (access points and controller
s).
Administrators SHOULD ensure that such devices are able to snoop DHCPv6-PD packe
ts, so the traffic from the delegated prefixes is not dropped.
</t>
<t>
It should be noted that using DHCPv6-PD makes it harder for an attacker to selec
t the spoofed source address.
When all clients are using the same shared link to form addresses, the attacker
might learn addresses used by other clients by listening to multicast Neighbor S
olicitations and Neighbour Advertisements.
In DHCPv6-PD environments, however, the attacker can only learn about other clie
nts's global addresses by listening to multicast DHCPv6 messages, which are not
transmitted so often, and may not be received by the client at all because they
are sent to multicast groups that are specific to DHCPv6 servers and relays.
</t>
</section> </section>
<section> <section>
<name>Migration Strategies and Co-existence with SLAAC Using Prefixes <name>Migration Strategies and Co-existence with SLAAC Using Prefixes from t
From PIO</name> he PIO</name>
<t>
It would be beneficial for the network to explicitly indicate <t>It would be beneficial for the network to explicitly indicate its
its support of DHCPv6-PD for connected clients. support of DHCPv6-PD for connected clients.</t>
</t> <ul>
<ul> <!-- [rfced] Is the closing parenthesis misplaced? If the suggested text is inc
<li> orrect, please clarify what happens when the prefix is too small.
In small networks (e.g., home networks), where the number of
clients is not too high, the number of available prefixes becomes a limiting fac Original:
tor. If every phone or laptop in a home network
If every phone or laptop in a home network were to re were to request a unique prefix suitable for SLAAC, the home
quest a unique prefix suitable for SLAAC, the home network might run out of pre network might run out of prefixes, if the prefix allocated to the
fixes, if the prefix allocated to the CPE by its ISP is too small (e.g., if an I CPE by its ISP is too small (e.g., if an ISP delegates a /60, it
SP delegates a /60, it would only be able to delegate 15 /64 prefixes to clients would only be able to delegate 15 /64 prefixes to clients).
).
So while the enterprise network administrator might w Perhaps:
ant all phones in the network to request a prefix, it would be highly undesirabl If every phone or laptop in a home network
e for the same phone to request a prefix when connecting to a home network. were to request a unique prefix suitable for SLAAC, the home
</li> network might run out of prefixes. If the prefix allocated to the
<li> Customer Premises Equipment (CPE) by its ISP is too small (e.g., if
When the network supports both a unique prefix per cl an ISP delegates a /60), it would only be able to delegate 15
ient and a PIO with A=1 as address assignment methods, it's highly desirable for /64 prefixes to clients.
the client NOT to use the PIO prefix to form global addresses and only use the -->
prefix delegated via DHCPv6-PD.
Starting both SLAAC using the PIO prefix and DHCPv6-P <li>In small networks (e.g., home networks), where the number of clients
D and deprecating that SLAAC addresses after receiving a delegated prefix would is not too high, the number of available prefixes becomes a limiting
be very disruptive for applications. factor. If every phone or laptop in a home network were to request a
If the client continues to use addresses formed from unique prefix suitable for SLAAC, the home network might run out of
the PIO prefix it would not only undermine the benefits of the proposed solution prefixes, if the prefix allocated to the Customer Premises Equipment (CPE)
(see <xref target="benefits"/>), but would also introduce complexity and unpred by its ISP is too small
ictability in the source address selection. (e.g., if an ISP delegates a /60, it would only be able to delegate 15
Therefore, the client needs to know what address assi /64 prefixes to clients). So while the enterprise network administrator
gnment method to use and whether to use the prefix in PIO or not, if the network might want all phones in the network to request a prefix, it would be
provides the PIO with A flag set. highly undesirable for the same phone to request a prefix when
</li> connecting to a home network.</li>
</ul>
<t> <!--[rfced] How may we update the text below for clarity? Specifically, what
The deployment model described in this document does not require two items does "both" refer to, and should "that SLAAC addresses" be plural or
the network to signal support of DHCPv6-PD: for example, devices acting as <xre singular?
f target="RFC7084"/> compatible routers will be able to receive prefixes via DHC
Pv6-PD even without such signalling. Also, some clients may decide to start DHCP Original:
v6-PD, and acquire prefixes, if they detect that the network does not provide ad
dresses via SLAAC. To fully achieve the benefits described in this section, <xre Starting both SLAAC using the PIO prefix and DHCPv6-PD and deprecating
f target="I-D.collink-6man-pio-pflag"/> defines a new PIO flag to signal that DH that SLAAC addresses after receiving a delegated prefix would be very
CPv6-PD is the preferred method of obtaining prefixes. disruptive for applications.
</t>
Perhaps:
Starting both DHCPv6-PD and SLAAC using the PIO prefix and then
deprecating that SLAAC address after receiving a delegated prefix
would be very disruptive for applications.
-->
<li>When the network supports both a unique prefix per client and a PIO
with A=1 as address assignment methods, it's highly desirable for the
client NOT to use the PIO prefix to form global addresses and instead only
use
the prefix delegated via DHCPv6-PD. Starting both SLAAC using the PIO
prefix and DHCPv6-PD and deprecating that SLAAC addresses after
receiving a delegated prefix would be very disruptive for applications.
If the client continues to use addresses formed from the PIO prefix, it
would not only undermine the benefits of the proposed solution (see
<xref target="benefits"/>), but it would also introduce complexity and
unpredictability in the source address selection. Therefore, the client
needs to know what address assignment method to use and whether or not to
use
the prefix in the PIO, if the network provides the PIO with the 'A' flag
set.</li>
</ul>
<t>The deployment model described in this document does not require the
network to signal support of DHCPv6-PD: for example, devices acting as
compatible routers <xref target="RFC7084"/> will be able to receive
prefixes via DHCPv6-PD even without such signaling. Also, some clients
may decide to start DHCPv6-PD and acquire prefixes if they detect that
the network does not provide addresses via SLAAC. To fully achieve the
benefits described in this section, <xref
target="I-D.ietf-6man-pio-pflag"/> defines a new PIO flag to signal
that DHCPv6-PD is the preferred method of obtaining prefixes.</t>
</section> </section>
<section anchor="benefits"> <section anchor="benefits">
<name>Benefits</name> <name>Benefits</name>
<t>
The proposed solution provides the following benefits: <!--[rfced] FYI - For ease of the reader, we have updated the text below as
</t> follows. Please review and let us know if this change alters your
intended meaning.
Original:
DHCPv6-PD logs and first-hop routers routing tables provide
complete information...
Current:
DHCPv6-PD logs and routing tables for first-hop routers provide
complete information...
-->
<t>The proposed solution provides the following benefits:</t>
<ul> <ul>
<li> <li>Network device resources (e.g., memory) need to scale to the
Network device resources (e.g., memory) need to scale number of devices, not the number of IPv6 addresses. The
to the number of devices, not the number of IPv6 addresses. first-hop routers have a single route per device pointing to the
The first-hop routers have a single route per device device's link-local address. This can potentially enable
pointing to the device's link-local address. This can potentially hardware cost savings; for example, if hardware such as wireless
enable hardware cost savings, for example if hardware LAN controllers is limited to supporting only a specific number
such as wireless LAN controllers is limited to supporting only of client addresses, or in VXLAN deployments where each client
a specific number of client addresses, or in VXLAN de address consumes one routing table entry.</li>
ployments where each client address consumes one routing table <li>The cost of having multiple addresses is offloaded to the
entry. clients. Hosts are free to create and use as many addresses as
</li> they need without imposing any additional costs onto the
<li> network.</li>
The cost of having multiple addresses is offloaded to <li>If all clients connected to the given link support this mode
the clients. of operation and can generate addresses from the delegated
Hosts are free to create and use as many addresses as prefixes, there is no reason to advertise a common prefix
they need without imposing any additional costs onto the network. assigned to that link in the PIO with the 'A' flag set. Therefore,
</li> it is
<li> possible to remove the global shared prefix from that link and
If all clients connected to the given link support th the router interface completely, so no global addresses are
is mode of operation and can generate addresses from the delegated prefixes, the on-link for the link. This would lead to reducing the attack
re is no reason to advertise a common prefix assigned to that link in PIO with ' surface for Neighbor Discovery attacks described in <xref
A' flag set. target="RFC6583"/>.</li>
Therefore it is possible to remove the global shared prefix from that link and t <li>DHCPv6-PD logs and routing tables for first-hop routers provide
he router interface completely, so no global addresses are on-link for the link. complete information on IPv6 to MAC mapping, which can be used
This would lead to reducing the attack surface for Neighbor Discovery attacks d for forensics and troubleshooting. Such information is much
escribed in <xref target="RFC6583"/>. less dynamic than the ND cache; therefore, it's much easier for an
</li> operator to collect and process it.</li>
<li> <li>A dedicated prefix per client allows the network
DHCPv6-PD logs and first-hop routers routing tables p administrator to create security policies per device (such as ACLs)
rovide complete information on IPv6 to MAC mapping, which can be used for forens even
ics and troubleshooting. if the client is using temporary addresses. This mitigates one
Such information is much less dynamic than ND cache a of the issues described in <xref target="I-D.ietf-opsec-ipv6-addres
nd therefore it's much easier for an operator to collect and process it. sing"/>.</li>
</li> <!-- [rfced] Does "or not" mean "not all of them work", "none of them work", or
<li> otherwise? Please clarify.
A dedicated prefix per client allows the network administrator to create per-dev
ice security policies (ACLs) even if the client is using temporary addresses. Th Original:
is mitigates one of the issues described in <xref target="I-D.ietf-opsec-ipv6-ad * Fate sharing: all global addresses used by a given client are
dressing"/>. routed as a single prefix. Either all of them work or not, which
</li> makes failures easier to diagnose and mitigate.
<li> -->
Fate sharing: all global addresses used by a given cl
ient are routed as a single prefix. <li>Fate sharing: all global addresses used by a given client
Either all of them work or not, which makes failures are routed as a single prefix. Either all of them work or not,
easier to diagnose and mitigate. which makes failures easier to diagnose and mitigate.</li>
</li> <li>Lower level of multicast traffic: less Neighbor Discovery
<li> <xref target="RFC4861"/> multicast packets, as the routers need to
Lower level of multicast traffic: less Neighbor Disco resolve only the clients' link-local addresses. Also, there is no Duplicate Ad
very (<xref target="RFC4861"/>) multicast packets, as there are only clients lin dress Detection (DAD) traffic
k-local addresses the routers need to resolve. Also, no DAD traffic except for c except for the clients' link-local addresses.</li>
lients' link-local addresses. <li>Ability to extend the network transparently. If the network
</li> delegates to the client a prefix of sufficient size to support
<li> SLAAC, the client can provide connectivity to other hosts, as
Ability to extend the network transparently. If the network delegates to the cli is possible in IPv4 with NAT (e.g., by acting as an IPv6 Customer E
ent a prefix of sufficient size to support SLAAC, the client can to provide conn dge (CE)
ectivity to other hosts, as is possible in IPv4 with NAT (e.g., by acting as an router as described in <xref target="RFC7084"/>).</li>
IPv6 CE router as described in <xref target="RFC7084"/>).
</li>
</ul> </ul>
</section> </section>
<section anchor="privacy"> <section anchor="privacy">
<name>Privacy Considerations</name> <name>Privacy Considerations</name>
<t> <t>If an eavesdropper or information collector is aware that a
If an eavesdropper or information collector is aware that a g given client is using the proposed mechanism, then they may be
iven client is using the able to track the client based on its prefix. The privacy
proposed mechanism, then they may be able to track the client implications of this are equivalent to the privacy implications of
based on its prefix. networks using stateful DHCPv6 address assignment: in both cases,
The privacy implications of this are equivalent to the privac the IPv6 addresses are determined by the server, either because
y implications of networks the server assigns a full 128-bit address in a shared prefix, or
using stateful DHCPv6 address assignment: in both cases, the because the server determines what prefix is delegated to the
IPv6 addresses are client. Administrators deploying the proposed mechanism can use
determined by the server, either because the server assigns a similar methods to mitigate the impact as the ones used today in
full 128-bit address in a networks that use stateful DHCPv6 address assignment.</t>
shared prefix, or because the server determines what prefix i
s delegated to the client. <t>Except for networks (such as datacenter networks) where hosts
Administrators deploying the proposed mechanism can use simil do not need temporary addresses <xref target="RFC8981"/>, the
ar methods to mitigate the network <bcp14>SHOULD</bcp14>:</t>
impact as the ones used today in networks that use stateful D
HCPv6 address assignment.</t>
<t>Except for networks (such as datacenter networks) where hosts do not
need temporary
addresses (<xref target="RFC8981"/>), the network SHOULD:</t>
<ul> <ul>
<li>Ensure that when a client requests a prefix, the prefix i <li>Ensure that when a client requests a prefix, the prefix is
s randomly assigned randomly assigned and not allocated deterministically.</li>
and not allocated deterministically.</li> <li>Use short prefix lifetimes (e.g., hours) to ensure that
<li>Use short prefix lifetimes (e.g., hours), to ensure that when a client disconnects and reconnects it gets a different
when a client prefix.</li>
disconnects and reconnects it gets a different prefix.</li> <li>Allow the client to have more than one prefix at the same
<li>Allow the client to have more than one prefix at the time. This allows the client to rotate prefixes using a
same time. This allows the client to rotate prefixes using a mecha mechanism similar to temporary addresses, but that operates on
nism similar to temporary prefixes instead of on individual addresses. In this case, the
addresses, but that operates on prefixes instead of on individual prefix's lifetime <bcp14>MUST</bcp14> be short enough to allow
addresses. the client to use a reasonable rotation interval without using
In this case the prefix's lifetime MUST be short enough to allow t too much address space. For example, if every 24 hours the
he client to use a client asks for a new prefix and stops renewing the old prefix,
reasonable rotation interval without using too much address space. and the Valid Lifetime of delegated prefixes is one hour, then
For example, if every 24 hours the the client asks for a new prefi the client will consume two prefixes for one hour out of 24
x and stops renewing the old hours, and thus will consume just under 1.05
prefix, and the Valid Lifetime of delegated prefixes is one hour, prefixes on average.</li>
then the client will consume </ul>
two prefixes for one hour out of 24 hours, and thus will on averag
e consume just under 1.05
prefixes.</li>
</ul>
</section> </section>
<section anchor="IANA"> <section anchor="IANA">
<!-- All drafts are required to have an IANA considerations section. See RFC 8126 for a guide.-->
<name>IANA Considerations</name> <name>IANA Considerations</name>
<t>This memo includes no request to IANA.</t> <t>This document has no IANA actions.</t>
</section> </section>
<section anchor="Security"> <section anchor="Security">
<!-- All drafts are required to have a security considerations section. Se e RFC 3552 for a guide. -->
<name>Security Considerations</name> <name>Security Considerations</name>
<t> A malicious (or just misbehaving) client might attempt to exhaust the <t>A malicious (or just misbehaving) client might attempt to exhaust the
DHCPv6-PD pool by sending a large number of requests with differing DUIDs. To pr DHCPv6-PD pool by sending a large number of requests with differing
event a misbehaving client from denying service to other clients, the DHCPv6 ser DHCP Unique Identifiers (DUIDs). To prevent a misbehaving client from deny
ver or relay MUST support limiting the number of prefixes delegated to a given c ing service to other
lient at any given time.</t> clients, the DHCPv6 server or relay <bcp14>MUST</bcp14> support limiting
<t>Networks can protect against malicious clients by authenticating dev the number of prefixes delegated to a given client at any given
ices using tokens that cannot be spoofed (e.g., 802.1x authentication) and limit time.</t>
ing the number of link-local addresses or MAC addresses that each client is allo
wed to use. Note that is not a new issue, as the same attack might be implemente <t>Networks can protect against malicious clients by authenticating
d using DHCPv4 or DHCPv6 IA_NA requests; in particular, while it is unlikely for devices using tokens that cannot be spoofed (e.g., 802.1x
clients to be able to exhaust an IA_NA address pool, clients using IA_NA can ex authentication) and limiting the number of link-local addresses or MAC
haust other resources such as DHCPv6 and routing infrastructure resources such s addresses that each client is allowed to use. Note that this is not a new
erver RAM, ND cache entries, TCAM entries, SAVI entries, etc. issue, as the same attack might be implemented using DHCPv4 or DHCPv6
</t> IA_NA requests; in particular, while it is unlikely for clients to be
<t> able to exhaust an IA_NA address pool, clients using IA_NA can exhaust
A malicious client might request a prefix and then release it very other resources such as DHCPv6 and routing infrastructure resources such a
quickly, causing routing convergence events on the relays. s
The impact of this attack can be reduced if the network rate-limits server RAM, ND cache entries, Ternary Content-Addressable Memory (TCAM) en
the amount of broadcast and multicast messages from the client. tries, SAVI entries, etc.</t>
</t>
<t> <t>A malicious client might request a prefix and then release it very
Delegating the same prefix for the same client introduces privacy c quickly, causing routing convergence events on the relays. The impact
oncerns. of this attack can be reduced if the network rate-limits the amount of
The proposed mitigation is discussed in <xref target="privacy"/>. broadcast and multicast messages from the client.</t>
</t>
<t> <t>Delegating the same prefix for the same client introduces privacy
Spoofing scenarios and prevention mechanisms are discussed in <xref concerns. The proposed mitigation is discussed in <xref
target="savi"/>. target="privacy"/>.</t>
</t>
<t>Spoofing scenarios and prevention mechanisms are discussed in <xref tar
get="savi"/>.</t>
</section> </section>
<!-- NOTE: The Acknowledgements and Contributors sections are at the end of this template -->
</middle> </middle>
<back> <back>
<displayreference target="I-D.ietf-6man-pio-pflag" to="PIO-PFLAG"/>
<displayreference target="I-D.ietf-dhc-rfc8415bis" to="RFC8415bis"/>
<displayreference target="I-D.ietf-dhc-addr-notification" to="ADDR-NOTIFICAT
ION"/>
<displayreference target="I-D.ietf-opsec-ipv6-addressing" to="IPv6-ADDRESS"/
>
<displayreference target="I-D.ietf-snac-simple" to="SNAC-SIMPLE"/>
<references> <references>
<name>References</name> <name>References</name>
<references> <references>
<name>Normative References</name> <name>Normative References</name>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.2 119.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.2 119.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.4 193.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.4 193.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 084.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 084.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.5 460.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.5 460.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.6 620.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.6 620.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.6 877.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.6 877.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 168.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 168.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 174.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 174.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 273.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 273.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 415.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 415.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 981.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 981.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 987.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 987.xml"/>
<!-- The recommended and simplest way to include a well known reference -->
</references> </references>
<references> <references>
<name>Informative References</name> <name>Informative References</name>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.3 704.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.3 704.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.4 861.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.4 861.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.4 862.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.4 862.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.6 459.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.6 459.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.6 583.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.6 583.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 039.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 039.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 278.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 278.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 348.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 348.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 421.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 421.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 513.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 513.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 653.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 653.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 934.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.7 934.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 200.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 200.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 501.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 501.xml"/>
<xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 504.xml"/> <xi:include href="https://www.rfc-editor.org/refs/bibxml/reference.RFC.8 504.xml"/>
<xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.
collink-6man-pio-pflag.xml"/> <!-- [I-D.ietf-6man-pio-pflag] IESG state: I-D Exists as of 07/09/24 -->
<xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.
ietf-6man-pio-pflag.xml"/>
<!-- [I-D.ietf-dhc-rfc8415bis] IESG state: I-D Exists as of 07/09/24 -->
<xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D. ietf-dhc-rfc8415bis.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D. ietf-dhc-rfc8415bis.xml"/>
<xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.i
etf-dhc-addr-notification"/> <!-- [I-D.ietf-dhc-addr-notification] IESG state: In RFC Ed Queue as of 07/09/24
-->
<reference anchor="I-D.ietf-dhc-addr-notification" target="https://datatracker.i
etf.org/doc/html/draft-ietf-dhc-addr-notification-13">
<front>
<title>Registering Self-generated IPv6 Addresses using DHCPv6</title>
<author fullname="Warren Kumari" initials="W." surname="Kumari">
<organization>Google, LLC</organization>
</author>
<author fullname="Suresh Krishnan" initials="S." surname="Krishnan">
<organization>Cisco Systems, Inc.</organization>
</author>
<author fullname="Rajiv Asati" initials="R." surname="Asati">
<organization>Independent</organization>
</author>
<author fullname="Lorenzo Colitti" initials="L." surname="Colitti">
<organization>Google, LLC</organization>
</author>
<author fullname="Jen Linkova" initials="J." surname="Linkova">
<organization>Google, LLC</organization>
</author>
<author fullname="Sheng Jiang" initials="S." surname="Jiang">
<organization>Beijing University of Posts and Telecommunications</organiza
tion>
</author>
<date day="16" month="May" year="2024"/>
</front>
<seriesInfo name="Internet-Draft" value="draft-ietf-dhc-addr-notification-13"/
>
</reference>
<!-- [I-D.ietf-opsec-ipv6-addressing] IESG state: Expired as of 04/08/24 -->
<xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D. ietf-opsec-ipv6-addressing.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D. ietf-opsec-ipv6-addressing.xml"/>
<!-- [I-D.ietf-snac-simple] IESG state: I-D Exists as of 06/20/24 -->
<xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D. ietf-snac-simple.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D. ietf-snac-simple.xml"/>
</references> </references>
</references> </references>
<section anchor="appendix" title="Appendix: Multiple Addresses Consideration <section anchor="appendix">
s"> <name>Multiple Addresses Considerations</name>
<t>While a typical IPv4 host normally has only one IPv4 address per interf
ace, an IPv6 device almost always has multiple addresses assigned to its interfa
ce.
At the very least, a host can be expected to have one link-local address, one te
mporary address, and, in most cases, one stable global address.
On a network providing NAT64 service, an IPv6-only host running the 464XLAT cust
omer-side translator (CLAT, <xref target="RFC6877"/>) would use a dedicated 464X
LAT address, configured via SLAAC (see Section 6.3 of <xref target="RFC6877"/>),
which brings the total number of addresses to 4.
Other common scenarios where the number of addresses per host interface might in
crease significantly, include but are not limited to:
</t>
<ul>
<li> <t>While a typical IPv4 host normally has only one IPv4 address per
Devices running containers/namespaces: each container/namespace would have multi interface, an IPv6 device almost always has multiple addresses assigned
ple addresses as described above. As a result, a device running just a few conta to its interface. At the very least, a host can be expected to have one
iners in a bridge mode can easily have 20 or more IPv6 addresses on the given li link-local address, one temporary address, and, in most cases, one
nk. stable global address. On a network providing NAT64 service, an
</li> IPv6-only host running the 464XLAT customer-side translator (CLAT) <xref
target="RFC6877"/> would use a dedicated 464XLAT address, configured
via SLAAC (see <xref target="RFC6877" section="6.3" sectionFormat="of"/>),
which brings
the total number of addresses to four. Other common scenarios where the
number of addresses per host interface might increase significantly
include but are not limited to:</t>
<li> <ul>
Networks assigning multiple prefixes to a given link: multihomed networks, netwo <li>Devices running containers/namespaces: each container/namespace
rks using ULA <xref target="RFC4193"/> and non-ULA prefixes together, or network would have multiple addresses as described above. As a result, a
performing a graceful renumbering from one prefix to another. device running just a few containers in a bridge mode can easily have
</li> 20 or more IPv6 addresses on the given link.</li>
<li>Networks assigning multiple prefixes to a given link: these include m
ultihomed
networks, networks using ULA <xref target="RFC4193"/> and non-ULA
prefixes together, or networks performing a graceful renumbering from
one prefix to another.</li>
</ul>
</ul> <t><xref target="RFC7934"/> discusses this aspect and explicitly states
that IPv6 deployments <bcp14>SHOULD NOT</bcp14> limit the number of IPv6
addresses a host can have. However, it has been observed that
networks often do limit the number of on-link addresses per device,
likely in an attempt to protect network resources and prevent DoS
attacks.</t>
<t> <t>The most common scenario of network-imposed limitations is ND proxy. M
<xref target="RFC7934"/> discusses this aspect and explicitly state any enterprise-scale wireless solutions
s that IPv6 deployments SHOULD NOT limit the number of IPv6 addresses a host can implement ND proxy to reduce the amount of broadcast and multicast
have. downstream (AP to clients) traffic and provide SAVI functions. To
However, it has been been observed that networks often do limit the perform ND proxy, a device usually maintains a table containing IPv6
number of on-link addresses per device, likely in an attempt to protect network and MAC addresses of connected clients. At least some implementations
resources and prevent DoS attacks. have hardcoded limits on how many IPv6 addresses per single MAC such a
</t> table can contain. When the limit is exceeded, the behavior is
<t> implementation dependent. Some vendors just fail to install an N+1 address
The most common scenario of network-imposed limitations is Neighbor D to the table. Others delete the oldest entry for this MAC and replace
iscovery (ND) proxy. it with the new address. In any case, the affected addresses lose
Many enterprise-scale wireless solutions implement ND proxy to redu network connectivity without receiving any implicit signal, with traffic
ce the amount of broadcast and multicast downstream (AP to clients) traffic and being silently dropped.</t>
provide SAVI functions.
To perform ND proxy, a device usually maintains a table, containing
IPv6 and MAC addresses of connected clients.
At least some implementations have hardcoded limits on how many IPv
6 addresses per single MAC such a table can contain.
When the limit is exceeded the behaviour is implementation-dependen
t. Some vendors just fail to install N+1 address to the table.
Others delete the oldest entry for this MAC and replace it with the
new address. In any case, the affected addresses lose network connectivity with
out receiving any implicit signal, with traffic being silently dropped.
</t>
</section> </section>
<section anchor="Acknowledgements" numbered="false"> <section anchor="Acknowledgements" numbered="false">
<!-- [REPLACE/DELETE] an Acknowledgements section is optional -->
<name>Acknowledgements</name> <name>Acknowledgements</name>
<t>Thanks to Harald Alvestrand, Nick Buraglio, Brian Carpenter, Tim Chown, <t>Thanks to <contact fullname="Harald Alvestrand"/>, <contact
Roman Danyliw, Gert Doering, David Farmer, Fernando Gont, Joel Halpern, Nick Hi fullname="Nick Buraglio"/>, <contact fullname="Brian Carpenter"/>,
lliard, Bob Hinden, Martin Hunek, Erik Kline, Warren Kumari, David Lamparter, An <contact fullname="Tim Chown"/>, <contact fullname="Roman Danyliw"/>,
drew McGregor, Tomek Mrugalski, Alexandre Petrescu, Jurgen Schonwalder, Pascal T <contact fullname="Gert Doering"/>, <contact fullname="David Farmer"/>,
hubert, Ole Troan, Eric Vyncke, Eduard Vasilenko, Timothy Winters, Chongfeng Xie <contact fullname="Fernando Gont"/>, <contact fullname="Joel Halpern"/>,
, Peter Yee for the discussions, their input, and all contributions.</t> <contact fullname="Nick Hilliard"/>, <contact fullname="Bob Hinden"/>,
<contact fullname="Martin Hunek"/>, <contact fullname="Erik Kline"/>,
<contact fullname="Warren Kumari"/>, <contact fullname="David
Lamparter"/>, <contact fullname="Andrew McGregor"/>, <contact
fullname="Tomek Mrugalski"/>, <contact fullname="Alexandre Petrescu"/>,
<contact fullname="Jurgen Schonwalder"/>, <contact fullname="Pascal
Thubert"/>, <contact fullname="Ole Troan"/>, <contact fullname="Eric
Vyncke"/>, <contact fullname="Eduard Vasilenko"/>, <contact
fullname="Timothy Winters"/>, <contact fullname="Chongfeng Xie"/>, and
<contact fullname="Peter Yee"/> for the discussions, their input, and all
contributions.</t>
</section> </section>
<section anchor="Contributors" numbered="false"> <!--[rfced] FYI - We have deleted the Contributors section of this document as
<!-- [REPLACE/DELETE] a Contributors section is optional --> it was empty. Please review. -->
<name>Contributors</name>
</section> <!--[rfced] Please review the following questions and changes regarding the
terminology used in this document.
a. Per Section 3.6 of RFC 7322 ("RFC Style Guide"), abbreviations must be
expanded upon first use. How may we expand AP, SSID, and ULA in the examples
below?
Originals:
When a client moves from between different attachment points on
the same link (e.g., roams between two APs while connected to the
same SSID or moves between two switchports belonging to the same
VLAN), the delegated prefix does not change
Many enterprise-scale wireless solutions
implement ND proxy to reduce the amount of broadcast and multicast
downstream (AP to clients) traffic and provide SAVI functions.
Networks assigning multiple prefixes to a given link: multihomed
networks, networks using ULA [RFC4193] and non-ULA prefixes
together, or network performing a graceful renumbering from one
prefix to another.
b. Per RFC 8415, IA_PD and IA_NA represent "Identity Association for Prefix
Delegation" and "Identity Association for Non-temporary Addresses",
respectively. Should these two terms be expanded and/or introduced in this
document's terminology list?
c. FYI - We have added expansions for abbreviations upon first use
per Section 3.6 of RFC 7322 ("RFC Style Guide"). Please review each
expansion in the document carefully to ensure correctness.
Local Internet Registry (LIR)
Customer Premises Equipment (CPE)
Virtual eXtensible Local Area Network (VXLAN)
DHCP Unique Identifiers (DUIDs)
Ternary Content-Addressable Memory (TCAM)
Duplicate Address Detection (DAD)
Customer Edge (CE)
interface identifier (IID) (per RFC 7421)
-->
<!--[rfced] Please review usage of the "/" character to separate terms in this
document (some instances included below). It may be unclear whether the
"/" stands for "or", "and", or "and/or". For clarity, please review and
let us know how/if these instances should be updated.
Originals:
Network devices need to maintain various types of
tables/hashes (Neighbor Cache on first-hop routers, Neighbor
Discovery Proxy caches on Layer 2 devices etc.).
However, providing a unique prefix per device is very uncommon in
enterprise-style networks, where nodes are usually connected to
broadcast segments/VLANs and each link has a single on-link prefix
assigned.
As most operators have some experience with IPv4, they can use a
similar approach for choosing the pool size and the timers (such as
T1/T2 timers).
Devices running containers/namespaces: each container/namespace would have
multiple addresses as described above.
-->
<!-- [rfced] Please review the "Inclusive Language" portion of the online
Style Guide <https://www.rfc-editor.org/styleguide/part2/#inclusive_language>
and let us know if any changes are needed. Note that updates may result in more
precise language that better convey the meaning to the readers.
For example, please consider whether "blackholing" should be updated in the
text below:
This leads to traffic blackholing and degraded customer experience.
-->
-&gt;
</back> </back>
</rfc> </rfc>
 End of changes. 87 change blocks. 
732 lines changed or deleted 1027 lines changed or added

This html diff was produced by rfcdiff 1.48.