<?xmlversion="1.0" encoding="US-ASCII"?>version='1.0' encoding='utf-8'?> <!DOCTYPE rfc SYSTEM"rfc2629.dtd" [ <!ENTITY RFC2119 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2119.xml"> <!ENTITY RFC2131 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2131.xml"> <!ENTITY RFC2132 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2132.xml"> <!ENTITY RFC2563 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2563.xml"> <!ENTITY RFC3927 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.3927.xml"> <!ENTITY RFC4039 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4039.xml"> <!ENTITY RFC4861 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4861.xml"> <!ENTITY RFC4957 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4957.xml"> <!ENTITY RFC6052 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6052.xml"> <!ENTITY RFC6146 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6146.xml"> <!ENTITY RFC6147 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6147.xml"> <!ENTITY RFC6877 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.6877.xml"> <!ENTITY RFC8174 SYSTEM "http://xml.resource.org/public/rfc/bibxml/reference.RFC.8174.xml"> ]> <?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?> <!-- used by XSLT processors --> <!-- For a complete list and description of processing instructions (PIs), please see http://xml.resource.org/authoring/README.html. --> <!-- Below are generally applicable Processing Instructions (PIs) that most I-Ds might want to use. (Here they are set differently than their defaults in xml2rfc v1.32) --> <?rfc strict="yes" ?> <!-- give errors regarding ID-nits and DTD validation --> <!-- control the table of contents (ToC) --> <?rfc toc="yes"?> <!-- generate a ToC --> <?rfc tocdepth="4"?> <!-- the number of levels of subsections in ToC. default: 3 --> <!-- control references --> <?rfc symrefs="yes"?> <!-- use symbolic references tags, i.e, [RFC2119] instead of [1] --> <?rfc sortrefs="yes" ?> <!-- sort the reference entries alphabetically --> <!-- control vertical white space (using these PIs as follows is recommended by the RFC Editor) --> <?rfc compact="yes" ?> <!-- do not start each main section on a new page --> <?rfc subcompact="no" ?> <!-- keep one blank line between list items --> <!-- end of list of popular I-D processing instructions -->"rfc2629-xhtml.ent"> <rfc xmlns:xi="http://www.w3.org/2001/XInclude" ipr="trust200902" updates="2563" obsoletes="" category="std"docName="draft-ietf-dhc-v6only-08"> <!-- category values: std, bcp, info, exp, and historic -->docName="draft-ietf-dhc-v6only-08" number="8925" submissionType="IETF" consensus="true" xml:lang="en" tocInclude="true" tocDepth="4" symRefs="true" sortRefs="true" version="3"> <!--***** FRONT MATTER *****xml2rfc v2v3 conversion 2.47.0 --> <front><!-- The abbreviated title is used in the page header - it is only necessary if the full title is longer than 39 characters --> <title>IPv6-Only-Preferred<title>IPv6-Only Preferred Option for DHCPv4</title><!-- add 'role="editor"' below for the editors if appropriate --><seriesInfo name="RFC" value="8925"/> <author fullname="Lorenzo Colitti" initials="L." surname="Colitti"> <organization>Google</organization> <address> <postal> <street>Shibuya 3-21-3</street><city>Shibuya</city> <region>Tokyo</region><region>Shibuya, Tokyo</region> <code>150-0002</code><country>JP</country><country>Japan</country> </postal><phone></phone><email>lorenzo@google.com</email> </address> </author> <author fullname="Jen Linkova" initials="J." surname="Linkova"> <organization>Google</organization> <address> <postal> <street>1 Darling Island Rd</street> <city>Pyrmont</city> <region>NSW</region> <code>2009</code><country>AU</country><country>Australia</country> </postal><phone></phone><email>furry@google.com</email> </address> </author> <author fullname="Michael C. Richardson" initials="M." surname="Richardson"> <organization abbrev="Sandelman">Sandelman Software Works</organization> <address> <email>mcr+ietf@sandelman.ca</email><uri>http://www.sandelman.ca/</uri><uri>https://www.sandelman.ca/</uri> </address> </author> <author fullname="Tomek Mrugalski" initials="T." surname="Mrugalski"> <organization abbrev="ISC">Internet Systems Consortium, Inc.</organization> <address> <postal><street>950 Charter Street</street> <city>Redwood City</city> <region>CA</region> <code>94063</code> <country>USA</country><street>PO Box 360</street> <city>Newmarket</city> <region>NH</region> <code>03857</code> <country>United States of America</country> </postal> <email>tomasz.mrugalski@gmail.com</email> </address> </author><date/> <!-- If the month and year are both specified and are the current ones, xml2rfc will fill in the current day for you. If only the current year is specified, xml2rfc will fill in the current day and month for you. If the year is not the current one, it is necessary to specify at least a month (xml2rfc assumes day="1" if not specified for the purpose of calculating the expiry date). With drafts it is normally sufficient to specify just the year. --> <!-- Meta-data Declarations --> <area>Internet</area> <workgroup>Dynamic Host Configuration</workgroup> <!-- WG name at the upperleft corner of the doc, IETF is fine for individual submissions. If this element is not present, the default is "Network Working Group", which is used by the RFC Editor as a nod to the history of the IETF. --> <keyword>template</keyword> <!-- Keywords will be incorporated into HTML output files in a meta tag but they have no effect on text or nroff output. If you submit your draft to the RFC Editor, the keywords will be used for the search engine. --><date month="October" year="2020"/> <abstract> <t> This document specifies a DHCPv4 option to indicate that a host supports an IPv6-only mode and is willing to forgo obtaining an IPv4 address if the network provides IPv6 connectivity. It also updatesRFC2563RFC 2563 to specifytheDHCPv4 server behavior when the server receives a DHCPDISCOVER not containing the Auto-Configure option but containing the new option defined in this document. </t> </abstract> </front> <middle> <sectiontitle="Introduction">numbered="true" toc="default"> <name>Introduction</name> <t> One of the biggest challenges of deploying IPv6-only LANs is that such networks might contain a rather heterogeneous collection of hosts. While some hosts are capable of operating in IPv6-only mode (either because the OS and all applications are IPv6-only capable or because the host has some form of 464XLAT <xreftarget="RFC6877"/>target="RFC6877" format="default"/> deployed), others might still have IPv4 dependencies and need IPv4 addresses to operate properly. To incrementallyrolloutroll out IPv6-only, network operators might need to provide IPv4 ondemanddemand, whereby a host receives an IPv4 address if it needs it, whileIPv6-only capableIPv6-only-capable hosts (such as modern mobile devices) are not allocated IPv4 addresses.TraditionallyTraditionally, that goal is achieved by placingIPv6-only capableIPv6-only-capable devicesintoin a dedicated IPv6-only network segment orWiFi SSID,Wi-Fi Service Set Identifier (SSID), while dual-stack devices reside in another network with IPv4 and DHCPv4 enabled.HoweverHowever, such an approach has a number of drawbacks,includingincluding, but not limitedto: <list style="symbols"> <t>to, the following: </t> <ul spacing="normal"> <li> Doubling the number of network segments leads to operational complexity and impact on performanceimpact,-- forinstanceinstance, due to high memory utilization caused by an increased number ofACLAccess Control List (ACL) entries.</t> <t></li> <li> Placing a hostintoin the correct network segment is problematic. For example, in the case of 802.11Wi-FiWi-Fi, the user might select the wrong SSID. In the case of wired 802.1xauthenticationauthentication, the authentication server might not have all the information required to make the correct decision and choose between an IPv6-only VLAN and a dual-stack VLAN.</t> </list> </t></li> </ul> <t> It would be beneficial for IPv6 deployment if operators could implement IPv6-mostly (or IPv4-on-demand) segments where IPv6-only hostsco-existcoexist with legacy dual-stack devices. The trivial solution of disabling the IPv4 stack onIPv6-only capableIPv6-only-capable hosts is notfeasiblefeasible, as those clients must be able to operate on IPv4-only networks as well. WhileIPv6-only capableIPv6-only-capable devices might use a heuristic approach to learning if the network provides IPv6-only functionality and stop using IPv4 if it does, such an approach might bepractically undesirable.undesirable in practice. One important reason is that when a host connects to a network, it does not knowifwhether the network is IPv4-only,dual-stackdual-stack, or IPv6-only. To ensure thattheconnectivity over whatever protocol is present becomes available as soon aspossiblepossible, the host usually starts configuring both IPv4 and IPv6 immediately. If hosts were to delay requesting IPv4 until IPv6 reachability is confirmed, that would penalize IPv4-only and dual-stack networks, which does not seem practical. Requesting IPv4 and then releasing it later, after IPv6 reachability is confirmed, might causeuser-visibleerrors that are visible to users, as it would be disruptive for applicationswhichthat have already started using the assigned IPv4address already. Insteadaddress. Instead, it would be useful to have a mechanismwhichthat would allow a host to indicate that its request for an IPv4 address is optional and a network to signal that IPv6-only functionality (such asNAT64,NAT64 <xreftarget="RFC6146"/>)target="RFC6146" format="default"/>) is available.The proposedThis document provides such a solutionis to introducevia a new DHCPv4 optionwhichthat a client uses to indicate that it does not need an IPv4 address if the network provides IPv6-only connectivity (as NAT64 and DNS64). If the particular network segment providesIPv4-on-demandIPv4 on demand, such clients would not be supplied with IPv4 addresses, while IPv4 addresses would be provided on IPv4-only or dual-stack segments without NAT64services IPv4 addresses will be provided.services. </t> <t> <xreftarget="RFC2563"/> introducestarget="RFC2563" format="default"/> introduced the Auto-Configure DHCPv4 option and describes DHCPv4serversserver behavior if no address is chosen for a host. This document updates <xreftarget="RFC2563"/>target="RFC2563" format="default"/> to modifytheserver behavior if the DHCPOFFER contains theIPv6-onlyIPv6-Only Preferred option. </t> <sectiontitle="Requirements Language">numbered="true" toc="default"> <name>Requirements Language</name> <t>The key words"MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY","<bcp14>MUST</bcp14>", "<bcp14>MUST NOT</bcp14>", "<bcp14>REQUIRED</bcp14>", "<bcp14>SHALL</bcp14>", "<bcp14>SHALL NOT</bcp14>", "<bcp14>SHOULD</bcp14>", "<bcp14>SHOULD NOT</bcp14>", "<bcp14>RECOMMENDED</bcp14>", "<bcp14>NOT RECOMMENDED</bcp14>", "<bcp14>MAY</bcp14>", and"OPTIONAL""<bcp14>OPTIONAL</bcp14>" in this document are to be interpreted as described inBCP 14BCP 14 <xref target="RFC2119"/> <xref target="RFC8174"/> when, and only when, they appear in all capitals, as shown here.</t> </section> <sectiontitle="Terminology"> <t> Dual-stacknumbered="true" toc="default"> <name>Terminology</name> <dl newline="false" spacing="normal"> <dt>Dual-stack network ordevice: adevice:</dt> <dd>A network or devicewhichthat has both versions of the Internet Protocol (IPv4 and IPv6) enabled andoperational. </t> <t> IPv6-only capable host: aoperational.</dd> <dt>IPv6-only-capable host:</dt><dd>A hostwhichthat does not require an IPv4 address and can operate on IPv6-only networks. More precisely, IPv6-only capability is specific to a given interface of the host: if someapplicationapplications on a host require IPv4 and the 464XLAT CLAT (customer-side translator) <xreftarget="RFC6877"/>target="RFC6877" format="default"/> is only enabled on one interface, the host is IPv6-only capable if connected to a NAT64 network via that interface. This document implies thatIPv6-only capableIPv6-only-capable hosts reach IPv4-only destinations via a NAT64 service provided by the network. <xref target="v6onlydef"/>format="default"/> discusses hypothetical scenariosoffor other transition technologies beingused. </t> <t> IPv4-requiring host: aused.</dd> <dt>IPv4-requiring host:</dt><dd>A hostwhichthat is not IPv6-only capable andcan notcannot operate in an IPv6-only network providing NAT64service. </t> <t> IPv4-on-demand: aservice.</dd> <dt>IPv4 on demand:</dt><dd>A deployment scenario where end hosts are expected to operate in IPv6-only mode by default and IPv4 addresses can be assigned to some hosts if those hosts explicitlyopt-inopt in toreceivingreceive IPv4 addresses.</t> <t> IPv6-mostly network: a</dd> <dt>IPv6-mostly network:</dt><dd>A networkwhichthat provides NAT64 (possibly with DNS64) service as well as IPv4 connectivity and allows the coexistence of IPv6-only,dual-stackdual-stack, and IPv4-only hosts on the same segment. Such a deployment scenario allows operators to incrementally turn off IPv4 on end hosts, while still providing IPv4 to deviceswhichthat require IPv4 to operate.But, IPv6-only capableBut IPv6-only-capable devices need not be assigned IPv4addresses. </t> <t> IPv6-only mode: aaddresses.</dd> <dt>IPv6-only mode:</dt><dd>A mode of operationwhenwhere a host acts as anIPv6-only capableIPv6-only-capable host and does not have IPv4 addresses assigned (except that IPv4 link-local addresses <xreftarget="RFC3927"/>target="RFC3927" format="default"/> may have beenconfigured). </t> <t> IPv6-only network: aconfigured).</dd> <dt>IPv6-only network:</dt><dd>A networkwhichthat does not provide routing functionality for IPv4 packets. Such networks may or may not allow intra-LAN IPv4 connectivity. An IPv6-only network usually provides access to IPv4-only resources via NAT64 <xreftarget="RFC6146"/>. </t> <t> NAT64: Networktarget="RFC6146" format="default"/>.</dd> <dt>NAT64:</dt><dd>Network Address and Protocol Translation from IPv6 Clients to IPv4 Servers <xreftarget="RFC6146"/>. </t> <t> RA: Router Advertisement, atarget="RFC6146" format="default"/>.</dd> <dt>Router Advertisement (RA):</dt><dd>A message used by IPv6 routers to advertise theirpresencepresence, together with various link and Internet parameters <xreftarget="RFC4861"/>. </t> <t> DNS64: atarget="RFC4861" format="default"/>.</dd> <dt>DNS64:</dt><dd>A mechanism for synthesizing AAAA records from A records <xreftarget="RFC6147"/>. </t> <t> Networktarget="RFC6147" format="default"/>.</dd> <dt>Network attachmentevent: A Link Upevent:</dt><dd>A link up event, as described by <xref target="RFC4957"/> whichformat="default"/>, that results in a host detecting an availablenetwork. </t> <t> Disablingnetwork.</dd> <dt>Disabling the IPv4 stack on the hostinterface: the hostinterface:</dt><dd> <t>Host behavior when thehost: <list style="symbols"> <t> doeshost</t> <ul> <li>does not send any IPv4 packets from thatinterface, </t> <t> dropsinterface,</li> <li>drops all IPv4 packets received on thatinterface and </t> <t> doesinterface, and</li> <li>does not forward any IPv4 packets to thatinterface. </t> </list> </t>interface.</li> </ul> </dd> </dl> </section> </section> <sectiontitle="Reasonsnumbered="true" toc="default"> <name>Reasons to Signal IPv6-Only Support in DHCPv4Packets">Packets</name> <t> For networkswhichthat contain a mix of bothIPv6-only capableIPv6-only-capable hosts and IPv4-requiringhosts,hosts andwhichthat utilize DHCPv4 for configuring the IPv4 network stack on hosts, it seems natural to leverage the same protocol to signal that IPv4 is discretional on a given segment. An ability to remotely disable IPv4 on a host can be seen as a new denial-of-service attack vector. Theproposedapproach provided in this document limits the attack surface to DHCPv4-related attacks without introducing new vulnerable elements. </t> <t> Another benefit of using DHCPv4 for signaling is that IPv4 will be disabled only if both the client and the server indicate IPv6-only capability. It allowsIPv6-only capableIPv6-only-capable hosts to turn off IPv4 only upon receiving an explicit signal from the network and operate in dual-stack or IPv4-only mode otherwise. In addition, theproposedmechanism defined in this document does not introduce any additional delays to the process of configuring an IP stack on hosts. If the network does not support IPv6-only/IPv4-on-demand mode, anIPv6-only capableIPv6-only-capable host would configure an IPv4 address as quickly asonany other host. </t> <t> Being a client/server protocol, DHCPv4 allows IPv4 to be selectively disabled on a per-host basis on a given network segment.CoexistenceThe coexistence of IPv6-only,dual-stackdual-stack, and even IPv4-only hosts on the same LAN would not only allow network administrators to preserve scarce IPv4 addresses but would also drastically simplify incremental deployment of IPv6-only networks, positively impacting IPv6 adoption. </t> </section> <sectiontitle="IPv6-Onlynumbered="true" toc="default"> <name>IPv6-Only PreferredOption">Option</name> <section anchor="Format"title="Option format">numbered="true" toc="default"> <name>Option Format</name> <figurealign="center" anchor="fig_Option" title="IPv6-Onlyanchor="fig_Option"> <name>IPv6-Only Preferred OptionFormat">Format</name> <artworkalign="center"><![CDATA[align="center" name="" type="" alt=""><![CDATA[ 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Code | Length | Value | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Value(contd)(cont.) |+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ]]></artwork>+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+]]></artwork> </figure> <t>Fields:</t><texttable style="none"> <ttcol></ttcol> <ttcol></ttcol> <c>Code: </c> <c> 8-bit<dl newline="false" spacing="normal"> <dt>Code:</dt> <dd>8-bit identifier of the IPv6-Only Preferred option code as assigned by IANA:TBD.108. The client includes the Code in the Parameter Request List in DHCPDISCOVER and DHCPREQUEST messages as described in <xreftarget="v4client"/>.</c> <c>Length:</c> <c> 8-bittarget="v4client" format="default"/>.</dd> <dt>Length:</dt> <dd>8-bit unsigned integer. The length of theoptionoption, excluding the Code and Length Fields. The serverMUST<bcp14>MUST</bcp14> set the length field to 4. The clientMUST<bcp14>MUST</bcp14> ignore the IPv6-Only Preferred option if the length field value is not4.</c> <c>Value:</c> <c> 32-bit4.</dd> <dt>Value:</dt> <dd><t>32-bit unsigned integer. The number of seconds for which the client should disable DHCPv4for(V6ONLY_WAIT configuration variable). If the server pool is explicitly configured with a V6ONLY_WAITtimertimer, the serverMUST<bcp14>MUST</bcp14> set the field to that configured value.OtherwiseOtherwise, the serverMUST<bcp14>MUST</bcp14> set it to zero. The clientMUST<bcp14>MUST</bcp14> process that field as described in <xreftarget="v4client"/>. Thetarget="v4client" format="default"/>.</t> <t>The client never sets thisfieldfield, as it never sends the full option but includes the option code in the Parameter Request List as described in <xreftarget="v4client"/>. </c> <c></c><c></c> </texttable>target="v4client" format="default"/>.</t></dd> </dl> </section> <section anchor="v4client"title="DHCPv4numbered="true" toc="default"> <name>DHCPv4 ClientBehavior">Behavior</name> <t> A DHCPv4 clientSHOULD<bcp14>SHOULD</bcp14> allow a device administrator to configure IPv6-onlypreferred modecapability either for a specific interface (to indicate that the device is IPv6-only capable if connected to a NAT64 network via that interface) or for all interfaces. If only a specific interface is configured as IPv6-onlycapablecapable, the DHCPv4 clientMUST NOT<bcp14>MUST NOT</bcp14> consider the hostto beanIPv6-only capableIPv6-only-capable host for the purpose of sending/receiving DHCPv4 packets over any other interfaces. </t> <t> The DHCPv4 client on an IPv4-requiring hostMUST NOT<bcp14>MUST NOT</bcp14> include theIPv6-onlyIPv6-Only Preferred option code in the Parameter Request List of any DHCPv4 packets andMUST<bcp14>MUST</bcp14> ignore that option in packets received from DHCPv4 servers. </t> <t> DHCPv4 clients running onIPv6-only capableIPv6-only-capable hostsSHOULD<bcp14>SHOULD</bcp14> include theIPv6-onlyIPv6-Only Preferred option code in the Parameter Request List in DHCPDISCOVER and DHCPREQUEST messages for interfaces so enabled and follow the processing as described below on aper enabled interfaceper-enabled-interface basis. </t> <t> If the client did not include theIPv6-onlyIPv6-Only Preferred option code in the Parameter Request Listoptionin the DHCPDISCOVER or DHCPREQUESTmessagemessage, itMUST<bcp14>MUST</bcp14> ignore theIPv6-onlyIPv6-Only Preferred option in any messages received from the server. </t> <t> If the client includes theIPv6-onlyIPv6-Only Preferred option code in the Parameter Request List and the DHCPOFFER message from the server contains a validIPv6-onlyIPv6-Only Preferred option, the clientSHOULD NOT<bcp14>SHOULD NOT</bcp14> request the IPv4 address provided in the DHCPOFFER. If theIPv6-onlyIPv6-Only Preferred option returned by the server contains a value greater than or equal to MIN_V6ONLY_WAIT, the clientSHOULD<bcp14>SHOULD</bcp14> set the V6ONLY_WAIT timer to that value. Otherwise, the clientSHOULD<bcp14>SHOULD</bcp14> set the V6ONLY_WAIT timer to MIN_V6ONLY_WAIT. The clientSHOULD<bcp14>SHOULD</bcp14> stop the DHCPv4 configuration process for V6ONLY_WAIT seconds or until a network attachment event, whichever happens first. The hostMAY<bcp14>MAY</bcp14> disable the IPv4 stack completely on the affected interface for V6ONLY_WAIT seconds or until the network attachment event, whichever happens first. </t> <t> TheIPv6-onlyIPv6-Only Preferred optionSHOULD<bcp14>SHOULD</bcp14> be included in the Parameter Request Listoptionin DHCPREQUEST messages (after receiving a DHCPOFFER without this option, foraan INIT-REBOOT, or when renewing or rebinding a leased address). If the DHCPv4 server responds with a DHCPACK that includes theIPv6-onlyIPv6-Only Preferred option, theclient behaviourclient's behavior depends on the client's state. If the client is in the INIT-REBOOTstatestate, itSHOULD<bcp14>SHOULD</bcp14> stop the DHCPv4 configuration process or disable the IPv4 stack completely for V6ONLY_WAIT seconds or until the network attachment event, whichever happens first. It alsoMAY<bcp14>MAY</bcp14> send a DHCPRELEASE message. If the client is in any otherstatestate, itSHOULD<bcp14>SHOULD</bcp14> continue to use the assigned IPv4 address until further DHCPv4 reconfiguration events. </t> <t> If the client includes theIPv6-onlyIPv6-Only Preferred option code in the Parameter Request List and the server responds with a DHCPOFFER message without a validIPv6-onlyIPv6-Only Preferred option, the clientMUST<bcp14>MUST</bcp14> proceed as normal with a DHCPREQUEST. </t> <t> If the client waits for multiple DHCPOFFER responses and selects one of them, itMUST<bcp14>MUST</bcp14> follow the processing for theIPv6-onlyIPv6-Only Preferred option based on the selected response. A clientMAY<bcp14>MAY</bcp14> use the presence of theIPv6-onlyIPv6-Only Preferred option as a selectioncriteria.criterion. </t> <t> When anIPv6-only capableIPv6-only-capable client receives the IPv6-Only Preferred option from the server, the clientMAY configurean<bcp14>MAY</bcp14> configure an IPv4 link-local address <xreftarget="RFC3927"/>.target="RFC3927" format="default"/>. In thatcase IPv6-only capablecase, IPv6-only-capable devices might still be able to communicate over IPv4 to other devices on the link. The Auto-ConfigureOptionoption <xreftarget="RFC2563"/>target="RFC2563" format="default"/> can be used to control the autoconfiguration of IPv4 link-localaddresses autoconfiguration.addresses. <xreftarget="autoconf"/>target="autoconf" format="default"/> discusses the interaction between theIPv6-onlyIPv6-Only Preferred option and the Auto-Configureoptions.option. </t> </section> <section anchor="v4srv"title="DHCPv4numbered="true" toc="default"> <name>DHCPv4 ServerBehavior">Behavior</name> <t> The DHCPv4 serverSHOULD<bcp14>SHOULD</bcp14> be able to configure all or individual pools to include theIPv6-only preferredIPv6-Only Preferred option in DHCPv4 responses if the client included the option code in the Parameter RequestList option.List. The DHCPv4 serverMAY<bcp14>MAY</bcp14> have a configuration option to specify the V6ONLY_WAIT timer for all or individual IPv6-mostly pools. </t> <t> The serverMUST NOT<bcp14>MUST NOT</bcp14> include theIPv6-onlyIPv6-Only Preferred option in the DHCPOFFER or DHCPACK message if theYIADDR field in the message does not belong to aselected pool is not configured as IPv6-mostly. The serverMUST NOT<bcp14>MUST NOT</bcp14> include theIPv6-onlyIPv6-Only Preferred option in the DHCPOFFER or DHCPACK message if the option was not present in the Parameter Request List sent by the client. </t> <t> If theIPv6-onlyIPv6-Only Preferred option is present in the Parameter Request List received from the client and the corresponding DHCPv4 pool is explicitly configured as belonging to an IPv6-mostly network segment, the serverMUST<bcp14>MUST</bcp14> include theIPv6-onlyIPv6-Only Preferred option when responding with the DHCPOFFER or DHCPACK message. If the server responds with theIPv6-onlyIPv6-Only Preferred option and the V6ONLY_WAIT timer is configured for the pool, the serverMUST<bcp14>MUST</bcp14> copy the configured value to theIPv6-onlyIPv6-Only Preferred option value field.OtherwiseOtherwise, itMUST<bcp14>MUST</bcp14> set the field to zero. The serverSHOULD NOT<bcp14>SHOULD NOT</bcp14> assign an address from the pool.InsteadInstead, itSHOULD<bcp14>SHOULD</bcp14> return 0.0.0.0 as the offered address. Alternatively, if offering 0.0.0.0 is notfeasible,feasible -- forexampleexample, due to some limitations of the server or the networkinfrastructure,infrastructure -- the serverMAY<bcp14>MAY</bcp14> include in the DHCPOFFER an available IPv4 address from thepool into the DHCPOFFERpool, as per recommendations in <xreftarget="RFC2131"/>.target="RFC2131" format="default"/>. In this case, the offered addressMUST<bcp14>MUST</bcp14> be a valid address that is not committed to any other client. Because the client is notexpectedever expected to request this address, the serverSHOULD NOT<bcp14>SHOULD NOT</bcp14> reserve the address andSHOULD NOT<bcp14>SHOULD NOT</bcp14> verify its uniqueness. If the client then issues a DHCPREQUEST for the address, the serverMUST<bcp14>MUST</bcp14> process it per <xreftarget="RFC2131"/>,target="RFC2131" format="default"/>, including replying with a DHCPACK for the address ifin the meantimeit has not been committed to anotherclient.client in the meantime. </t> <t> If a client includes both aRapid-CommitRapid Commit option <xreftarget="RFC4039"/>target="RFC4039" format="default"/> and an IPv6-Only Preferred option in the DHCPDISCOVERmessagemessage, the serverSHOULD NOT<bcp14>SHOULD NOT</bcp14> honor theRapid-CommitRapid Commit option if the response to the client would contain theIPv6-onlyIPv6-Only Preferredoption to the client.option. ItSHOULD<bcp14>SHOULD</bcp14> instead respond with a DHCPOFFER as indicated above. </t> <t> If the server receives a DHCPREQUEST containing theIPv6-onlyIPv6-Only Preferred option for the address from a pool configured as IPv6-mostly, the serverMUST<bcp14>MUST</bcp14> process it per <xreftarget="RFC2131"/>.target="RFC2131" format="default"/>. </t> <section anchor="autoconf"title="Interactionnumbered="true" toc="default"> <name>Interaction withRFC2563">RFC 2563</name> <t> <xreftarget="RFC2563"/>target="RFC2563" format="default"/> defines an Auto-Configure DHCPv4 option to disable IPv4 link-local address configuration for IPv4 clients. Clients can supportboth, neither or just one ofboth the IPv6-Only Preferred option and the Auto-Configureoptions.option, just one of the options, or neither option. If a client sends both the IPv6-Only Preferred option and the Auto-Configureoptionsoption, the network administrator can prevent the host from configuring an IPv4 link-local address on an IPv6-mostly network. To achievethisthis, the server needs to send a DHCPOFFERwhichthat contains a 'yiaddr' of0x00000000,0.0.0.0, and the Auto-Configure flagsayingset to "DoNotAutoConfigure". </t> <t>HoweverHowever, special care should be taken in a situationwhenwhere a server supports both options and receives just an IPv6-Only Preferred option from a client.Section 2.3 of<xreftarget="RFC2563"/>target="RFC2563" sectionFormat="of" section="2.3"/> states that if no address is chosen for the host (which would be the case forIPv6-only capableIPv6-only-capable clients on an IPv6-mostlynetwork) then:network), then "If the DHCPDISCOVER does not contain the Auto-Configure option, it is not answered." Such behavior would be undesirable for clients supporting the IPv6-Only Preferred option without supporting the Auto-Configureoptionoption, as they would not receive any response from the server and would keepasking,requesting a response instead of disabling DHCPv4 for V6ONLY_WAIT seconds.ThereforeTherefore, the following update is made toSection 2.3 of<xreftarget="RFC2563"/>"target="RFC2563" sectionFormat="of" section="2.3"/>. </t> <t> OLD TEXT: </t><t> --- </t> <t> However,<blockquote><t>However, if no address is chosen for the host, a few additional stepsMUST<bcp14>MUST</bcp14> betaken. </t>taken.</t> <t> If the DHCPDISCOVER does not contain the Auto-Configure option, it is not answered. </t><t> --- </t></blockquote> <t> NEW TEXT: </t><t> --- </t> <t> However,<blockquote><t>However, if no address is chosen for the host, a few additional stepsMUST<bcp14>MUST</bcp14> be taken. </t> <t> If the DHCPDISCOVER does not contain the Auto-Configure option and the IPv6-Only Preferred option is not present, it is not answered. If the DHCPDISCOVER does not contain the Auto-Configure option but contains the IPv6-Only Preferred option, the processing rules for the IPv6-Only Preferred option apply. </t><t> --- </t></blockquote> </section> </section> <section anchor="vars"title="Constantsnumbered="true" toc="default"> <name>Constants and ConfigurationVariables"> <texttable style="none"> <ttcol></ttcol> <ttcol></ttcol> <c>V6ONLY_WAIT</c> <c>TheVariables</name> <dl newline="false" spacing="normal"> <dt>V6ONLY_WAIT:</dt> <dd>The time for which the clientSHOULD<bcp14>SHOULD</bcp14> stop the DHCPv4 configuration process. The valueMUST NOT<bcp14>MUST NOT</bcp14> be less than MIN_V6ONLY_WAIT seconds. Default: 1800seconds</c> <c>MIN_V6ONLY_WAIT</c> <c>Theseconds</dd> <dt>MIN_V6ONLY_WAIT:</dt> <dd>The lower boundary for V6ONLY_WAIT. Value: 300seconds</c> <c></c><c></c> </texttable>seconds</dd> </dl> </section> </section> <section anchor="v6onlydef"title="IPv6-Onlynumbered="true" toc="default"> <name>IPv6-Only TransitionTechnologies Considerations">Technology Considerations</name> <t> Until IPv6 adoption in the Internet reaches 100%, communication between an IPv6-only host and an IPv4-only destination requires some form of a transition mechanism deployed in the network. At the time of writing, the only such mechanism that is widely supported by end hosts is NAT64 <xreftarget="RFC6146"/>target="RFC6146" format="default"/> (either with or without 464XLAT).ThereforeTherefore, theIPv6-onlyIPv6-Only Preferred option is only sent by hosts capable of operating on NAT64 networks. In a typical deployment scenario, a network administrator would not configure the DHCPv4 server to return theIPv6-onlyIPv6-Only Preferred option unless the network provides NAT64 service. </t> <t> Hypothetically, it is possible for multiple transition technologies to coexist. In suchscenarioa scenario, some form of negotiation would be required between a client and a server to ensure that the transition technology supported by the client is the one the network provides.HoweverHowever, it seems unlikely that any new transition technology would arise and be widely adopted inanythe foreseeable future.ThereforeTherefore, adding support for non-existing technologies seems to besuboptimal and the proposed mechanismsuboptimal, so this document implies that NAT64 is used to facilitate connectivity between IPv6 and IPv4. In the unlikely event that a new transition mechanism becomes widely deployed, the applicability of theIPv6-Only-PreferredIPv6-Only Preferred option to that mechanism will depend on the nature of the new mechanism. If the new mechanism is designed in such a way that it's fully transparent for hosts that support NAT64 and theIPv6-Only-PreferredIPv6-Only Preferred option, then the option can continue to be used with the new mechanism. If the new mechanism is not compatible withNAT64,NAT64 and implementation on the host side is required to support it, then a new DHCPv4 option needs to be defined. </t> <t> It shouldbealso be noted that declaring a host (technically, a host interface) IPv6-only capable is a policy decision. For example,<list style="symbols"> <t></t> <ul spacing="normal"> <li> Anoperating systemOS vendor may make such a decision and configure their DHCPv4 clients to send the IPv6-Only Preferred option by default if the OS has a 464XLAT CLAT <xreftarget="RFC6877"/>target="RFC6877" format="default"/> enabled.</t> <t></li> <li> An enterprise network administrator may provision the corporate hosts as IPv6-only capable if all applications that users are supposed to run have been tested in an IPv6-only environment (or if a 464XLAT CLAT is enabled on the devices).</t> <t> IoT</li> <li> Internet of Things (IoT) devices may be shipped inIPv6-only capableIPv6-only-capable mode if they are designed to connect to an IPv6-enabled cloud destination only.</t> </list> </t></li> </ul> </section> <section anchor="IANA"title="IANA Considerations">numbered="true" toc="default"> <name>IANA Considerations</name> <t>The IANAis requested to assignhas assigned a new DHCPv4Optionoption code for the IPv6-Only Preferred option from theBOOTP"BOOTP Vendor Extensions andDHCPv4 OptionsDHCP Options" registry, located athttps://www.iana.org/assignments/bootp-dhcp-parameters/bootp-dhcp-parameters.xhtml#options . If possible, please assign option code 108. </t> <texttable anchor="option_table"> <ttcol align="left">Tag</ttcol> <ttcol align="left">Name</ttcol> <ttcol align="left">Data Length</ttcol> <ttcol align="left">Meaning</ttcol> <ttcol align="left">Reference</ttcol> <c>TBD (proposed value: 108)</c> <c>IPv6-only Preferred option</c> <c>4</c> <c>Number<eref target="https://www.iana.org/assignments/bootp-dhcp-parameters/" brackets="angle"/>. </t> <dl newline="false" spacing="compact"> <dt>Tag:</dt> <dd>108</dd> <dt>Name:</dt> <dd>IPv6-Only Preferred</dd> <dt>Data Length:</dt> <dd>4</dd> <dt>Meaning:</dt> <dd>Number of secondsto disablethat DHCPv4for</c> <c>draft-ietf-dhc-v6only</c> </texttable>should be disabled</dd> <dt>Reference:</dt> <dd>RFC 8925</dd> </dl> </section> <section anchor="Security"title="Security Considerations">numbered="true" toc="default"> <name>Security Considerations</name> <t> An attacker might send a spoofed DHCPOFFER containingIPv6-onlyan IPv6-Only Preferred option with the value field set to a large number, such as 0xffffffff, effectively disabling DHCPv4 on clients supporting the option. If the network isIPv4-onlyIPv4-only, such clients would lose connectivity, while on a dual-stack network without NAT64serviceservice, only connectivity to IPv4-only destinations would be affected.The recoveryRecovery from such an attack would require triggering a network attachment event.HoweverHowever, it should be noted that if the network does not provide protection from a rogue DHCPv4serverserver, the similar attack vector can be executed by offering an invalid address and setting the Lease Time option <xref target="RFC2132"/> value field to 0xffffffff. The latter attack would affect allhosts,hosts -- not just hosts that support theIPv6-onlyIPv6-Only Preferred option.ThereforeTherefore, the security measures against rogue DHCPv4 servers would be sufficient to preventtheattacks specific toIPv6-onlythe IPv6-Only Preferred option.AdditionallyAdditionally, such attacks can only be executed if the victim prefers the rogue DHCPOFFER overthelegitimateones. Thereforeoffers. Therefore, for the attack to besuccessfulsuccessful, the attacker needs to know the selection criteria used by the client andtobe able to make its rogue offermore preferable.preferable to other offers. </t> <t> It should be noted that disabling IPv4 on a host upon receiving theIPv6-onlyIPv6-Only Preferred option from the DHCPv4 server protects the host from IPv4-related attacks and therefore could be considered a securityfeaturefeature, as it reduces the attack surface. </t> </section> </middle> <back> <references> <name>References</name> <references> <name>Normative References</name> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2119.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2131.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2563.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.3927.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.4039.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8174.xml"/> </references> <references> <name>Informative References</name> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2132.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.4861.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.4957.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6146.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6147.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6877.xml"/> </references> </references> <section anchor="Acknowledgements"title="Acknowledgements">numbered="false" toc="default"> <name>Acknowledgements</name> <t> Thanks to the following people (in alphabetical order) for their review and feedback:Mohamed Boucadair, Martin Duke, Russ Housley, Sheng Jiang, Benjamin Kaduk, Murray Kucherawy, Ted Lemon, Roy Marples, Bjorn Mork, Alvaro Retana, Peng Shuping, Pascal Thubert, Bernie Volz, Eric Vyncke, Robert Wilton. Authors<contact fullname="Mohamed Boucadair"/>, <contact fullname="Martin Duke"/>, <contact fullname="Russ Housley"/>, <contact fullname="Sheng Jiang"/>, <contact fullname="Benjamin Kaduk"/>, <contact fullname="Murray Kucherawy"/>, <contact fullname="Ted Lemon"/>, <contact fullname="Roy Marples"/>, <contact fullname="Bjorn Mork"/>, <contact fullname="Alvaro Retana"/>, <contact fullname="Peng Shuping"/>, <contact fullname="Pascal Thubert"/>, <contact fullname="Bernie Volz"/>, <contact fullname="Éric Vyncke"/>, and <contact fullname="Robert Wilton"/>. The authors would like to thankBob Hinden and Brian Carpenter<contact fullname="Bob Hinden"/> and <contact fullname="Brian Carpenter"/> for the initial idea of signaling IPv6-only capability to hosts. Special thanks toErik Kline, Mark Townsley and Maciej Zenczykowski<contact fullname="Erik Kline"/>, <contact fullname="Mark Townsley"/>, and <contact fullname="Maciej Zenczykowski"/> for the discussionwhichthat led to the idea ofsignallingsignaling IPv6-only capability over DHCPv4. </t> </section></middle> <!-- *****BACK MATTER ***** --> <back> <references title="Normative References"> &RFC2119; &RFC2131; &RFC2563; &RFC3927; &RFC4039; &RFC8174; </references> <references title="Informative References"> <!-- &RFC6052; --> &RFC4861; &RFC4957; &RFC6146; &RFC6147; &RFC6877; </references></back> </rfc>