Independent Submission F. Templin, Ed.

Independent Submission F. Templin, Ed.
Request for Comments: 5558 Boeing Research & Technology
Category: Informational February 2010
ISSN: 2070-1721

Virtual Enterprise Traversal (VET)

Abstract

Enterprise networks connect routers over various link types, and may
also connect to provider networks and/or the global Internet.
Enterprise network nodes require a means to automatically provision
IP addresses/prefixes and support internetworking operation in a wide
variety of use cases including Small Office, Home Office (SOHO)
networks, Mobile Ad hoc Networks (MANETs), multi-organizational
corporate networks and the interdomain core of the global Internet
itself. This document specifies a Virtual Enterprise Traversal (VET)
abstraction for autoconfiguration and operation of nodes in
enterprise networks.

Status of This Memo

This document is not an Internet Standards Track specification; it is
published for informational purposes.

This is a contribution to the RFC Series, independently of any other
RFC stream. The RFC Editor has chosen to publish this document at
its discretion and makes no statement about its value for
implementation or deployment. Documents approved for publication by
the RFC Editor are not a candidate for any level of Internet
Standard; see Section 2 of RFC 5741.

Information about the current status of this document, any errata,
and how to provide feedback on it may be obtained at
http://www.rfc-editor.org/info/rfc5558.

Templin Informational [Page 1]

RFC 5558 VET February 2010

IESG Note

This RFC is not a candidate for any level of Internet Standard. The
IETF disclaims any knowledge of the fitness of this RFC for any
purpose and in particular notes that the decision to publish is not
based on IETF review for such things as security, congestion control,
or inappropriate interaction with deployed protocols. The RFC Editor
has chosen to publish this document at its discretion. Readers of
this RFC should exercise caution in evaluating its value for
implementation and deployment. See RFC 3932 for more information.

Note that the IETF AUTOCONF Working Group is working on a similar
protocol solution that may become available in the future.

Copyright Notice

Copyright (c) 2010 IETF Trust and the persons identified as the
document authors. All rights reserved.

This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document.

Templin Informational [Page 2]

RFC 5558 VET February 2010

Table of Contents

1. Introduction ....................................................4
2. Terminology .....................................................6
3. Enterprise Characteristics .....................................10
4. Autoconfiguration ..............................................11
4.1. Enterprise Router (ER) Autoconfiguration ..................12
4.2. Enterprise Border Router (EBR) Autoconfiguration ..........13
4.2.1. VET Interface Autoconfiguration ....................13
4.2.1.1. Interface Initialization ..................14
4.2.1.2. Enterprise Border Gateway
Discovery and Enterprise Identification ...14
4.2.1.3. EID Configuration .........................15
4.2.2. Provider-Aggregated (PA) EID Prefix
Autoconfiguration ..................................15
4.2.3. Provider-Independent (PI) EID Prefix
Autoconfiguration ..................................16
4.3. Enterprise Border Gateway (EBG) Autoconfiguration .........17
4.4. VET Host Autoconfiguration ................................17
5. Internetworking Operation ......................................18
5.1. Routing Protocol Participation ............................18
5.2. RLOC-Based Communications .................................18
5.3. EID-Based Communications ..................................18
5.4. IPv6 Router Discovery and Prefix Registration .............18
5.4.1. IPv6 Router and Prefix Discovery ...................18
5.4.2. IPv6 PA Prefix Registration ........................19
5.4.3. IPv6 PI Prefix Registration ........................20
5.4.4. IPv6 Next-Hop EBR Discovery ........................21
5.5. IPv4 Router Discovery and Prefix Registration .............23
5.6. VET Encapsulation .........................................24
5.7. SEAL Encapsulation ........................................24
5.8. Generating Errors .........................................25
5.9. Processing Errors .........................................25
5.10. Mobility and Multihoming Considerations ..................26
5.11. Multicast ................................................27
5.12. Service Discovery ........................................28
5.13. Enterprise Partitioning ..................................29
5.14. EBG Prefix State Recovery ................................29
6. Security Considerations ........................................30
7. Related Work ...................................................30
8. Acknowledgements ...............................................31
9. Contributors ...................................................31
10. References ....................................................31
10.1. Normative References .....................................31
10.2. Informative References ...................................33
Appendix A. Duplicate Address Detection (DAD) Considerations .... 36

Templin Informational [Page 3]

RFC 5558 VET February 2010

1. Introduction

Enterprise networks [RFC4852] connect routers over various link types
(see [RFC4861], Section 2.2). The term "enterprise network" in this
context extends to a wide variety of use cases and deployment
scenarios. For example, an "enterprise" can be as small as a SOHO
network, as complex as a multi-organizational corporation, or as
large as the global Internet itself. Mobile Ad hoc Networks (MANETs)
[RFC2501] can also be considered as a challenging example of an
enterprise network, in that their topologies may change dynamically
over time and that they may employ little/no active management by a
centralized network administrative authority. These specialized
characteristics for MANETs require careful consideration, but the
same principles apply equally to other enterprise network scenarios.

This document specifies a Virtual Enterprise Traversal (VET)
abstraction for autoconfiguration and internetworking operation,
where addresses of different scopes may be assigned on various types
of interfaces with diverse properties. Both IPv4 [RFC0791] and IPv6
[RFC2460] are discussed within this context. The use of standard
DHCP [RFC2131] [RFC3315] and neighbor discovery [RFC0826] [RFC1256]
[RFC4861] mechanisms is assumed unless otherwise specified.

Templin Informational [Page 4]

RFC 5558 VET February 2010

Provider-Edge Interfaces
x x x
| | |
+--------------------+---+--------+----------+ E
| | | | | n
| I | | .... | | t
| n +---+---+--------+---+ | e
| t | +--------+ /| | r
| e I x----+ | Host | I /*+------+--< p I
| r n | |Function| n|**| | r n
| n t | +--------+ t|**| | i t
| a e x----+ V e|**+------+--< s e
| l r . | E r|**| . | e r
| f . | T f|**| . | f
| V a . | +--------+ a|**| . | I a
| i c . | | Router | c|**| . | n c
| r e x----+ |Function| e \*+------+--< t e
| t s | +--------+ \| | e s
| u +---+---+--------+---+ | r
| a | | .... | | i
| l | | | | o
+--------------------+---+--------+----------+ r
| | |
x x x
Enterprise-Edge Interfaces

Figure 1: Enterprise Router (ER) Architecture

Figure 1 above depicts the architectural model for an Enterprise
Router (ER). As shown in the figure, an ER may have a variety of
interface types including enterprise-edge, enterprise-interior,
provider-edge, internal-virtual, as well as VET interfaces used for
IP-in-IP encapsulation. The different types of interfaces are
defined, and the autoconfiguration mechanisms used for each type are
specified. This architecture applies equally for MANET routers, in
which enterprise-interior interfaces correspond to the wireless
multihop radio interfaces typically associated with MANETs. Out of
scope for this document is the autoconfiguration of provider
interfaces, which must be coordinated in a manner specific to the
service provider's network.

Enterprise networks must have a means for supporting both Provider-
Independent (PI) and Provider-Aggregated (PA) IP prefixes. This is
especially true for enterprise scenarios that involve mobility and
multihoming. Also in scope are ingress filtering for multihomed
sites, adaptation based on authenticated ICMP feedback from on-path
routers, effective tunnel path MTU mitigations, and routing scaling
suppression as required in many enterprise network scenarios.

Templin Informational [Page 5]

RFC 5558 VET February 2010

Recognizing that one size does not fit all, the VET specification
provides adaptable mechanisms that address these issues, and more, in
a wide variety of enterprise network use cases.

VET represents a functional superset of 6over4 [RFC2529] and Intra-
Site Automatic Tunnel Addressing Protocol (ISATAP) [RFC5214], and it
further supports additional encapsulations such as IPsec [RFC4301],
Subnetwork Encapsulation and Adaptation Layer (SEAL) [RFC5320], etc.
Together, these technologies serve as functional building blocks for
a new Internetworking architecture known as Routing and Addressing in
Networks with Global Enterprise Recursion [RFC5720][RANGERS].

The VET principles can be either directly or indirectly traced to the
deliberations of the ROAD group in January 1992, and also to still
earlier works including NIMROD [RFC1753], the Catenet model for
internetworking [CATENET] [IEN48] [RFC2775], etc. [RFC1955] captures
the high-level architectural aspects of the ROAD group deliberations
in a "New Scheme for Internet Routing and Addressing (ENCAPS) for
IPNG".

VET is related to the present-day activities of the IETF AUTOCONF,
DHC, IPv6, MANET, and v6OPS working groups, as well as the IRTF RRG
working group.

2. Terminology

The mechanisms within this document build upon the fundamental
principles of IP-in-IP encapsulation. The terms "inner" and "outer"
are used to, respectively, refer to the innermost IP {address,
protocol, header, packet, etc.} *before* encapsulation, and the
outermost IP {address, protocol, header, packet, etc.} *after*
encapsulation. VET also allows for inclusion of "mid-layer"
encapsulations between the inner and outer layers, including IPsec
[RFC4301], the Subnetwork Encapsulation and Adaptation Layer (SEAL)
[RFC5320], etc.

The terminology in the normative references apply; the following
terms are defined within the scope of this document:

subnetwork
the same as defined in [RFC3819].

enterprise
the same as defined in [RFC4852]. An enterprise is also
understood to refer to a cooperative networked collective with a
commonality of business, social, political, etc. interests.

Templin Informational [Page 6]

RFC 5558 VET February 2010

Minimally, the only commonality of interest in some enterprise
network scenarios may be the cooperative provisioning of
connectivity itself.

site
a logical and/or physical grouping of interfaces that connect a
topological area less than or equal to an enterprise in scope. A
site within an enterprise can, in some sense, be considered as an
enterprise unto itself.

Mobile Ad hoc Network (MANET)
a connected topology of mobile or fixed routers that maintain a
routing structure among themselves over dynamic links, where a
wide variety of MANETs share common properties with enterprise
networks. The characteristics of MANETs are defined in [RFC2501],
Section 3.

enterprise/site/MANET
throughout the remainder of this document, the term "enterprise"
is used to collectively refer to any of enterprise/site/MANET,
i.e., the VET mechanisms and operational principles can be applied
to enterprises, sites, and MANETs of any size or shape.

Enterprise Router (ER)
As depicted in Figure 1, an Enterprise Router (ER) is a fixed or
mobile router that comprises a router function, a host function,
one or more enterprise-interior interfaces, and zero or more
internal virtual, enterprise-edge, provider-edge, and VET
interfaces. At a minimum, an ER forwards outer IP packets over
one or more sets of enterprise-interior interfaces, where each set
connects to a distinct enterprise.

Enterprise Border Router (EBR)
an ER that connects edge networks to the enterprise and/or
connects multiple enterprises together. An EBR is a tunnel
endpoint router, and it configures a separate VET interface over
each set of enterprise-interior interfaces that connect the EBR to
each distinct enterprise. In particular, an EBR may configure
multiple VET interfaces -- one for each distinct enterprise. All
EBRs are also ERs.

Enterprise Border Gateway (EBG)
an EBR that connects VET interfaces configured over child
enterprises to a provider network -- either directly via a
provider-edge interface or indirectly via another VET interface
configured over a parent enterprise. EBRs may act as EBGs on some
VET interfaces and as ordinary EBRs on other VET interfaces. All
EBGs are also EBRs.

Templin Informational [Page 7]

RFC 5558 VET February 2010

enterprise-interior interface
an ER's attachment to a link within an enterprise. Packets sent
over enterprise-interior interfaces may be forwarded over multiple
additional enterprise-interior interfaces within the enterprise
before they are forwarded via an enterprise-edge interface,
provider-edge interface, or a VET interface configured over a
different enterprise. Enterprise-interior interfaces connect
laterally within the IP network hierarchy.

enterprise-edge interface
an EBR's attachment to a link (e.g., an Ethernet, a wireless
personal area network, etc.) on an arbitrarily complex edge
network that the EBR connects to an enterprise and/or provider
network. Enterprise-edge interfaces connect to lower levels
within the IP network hierarchy.

provider-edge interface
an EBR's attachment to the Internet or to a provider network
outside of the enterprise via which the Internet can be reached.
Provider-edge interfaces connect to higher levels within the IP
network hierarchy.

internal-virtual interface
an interface that is internal to an EBR and does not in itself
directly attach to a tangible physical link, e.g., an Ethernet
cable. Examples include a loopback interface, a virtual LAN
interface, or some form of tunnel interface.

Virtual Enterprise Traversal (VET)
an abstraction that uses IP-in-IP encapsulation to create an
overlay that spans an enterprise in a single (inner) IP hop.

VET interface
an EBR's tunnel virtual interface used for Virtual Enterprise
Traversal. The EBR configures a VET interface over a set of
underlying interfaces belonging to the same enterprise. When
there are multiple distinct enterprises (each with their own
distinct set of underlying interfaces), the EBR configures a
separate VET interface over each set of underlying interfaces,
i.e., the EBR configures multiple VET interfaces.

The VET interface encapsulates each inner IP packet in any mid-
layer headers plus an outer IP header, then it forwards it on an
underlying interface such that the Time to Live (TTL) / Hop Limit
in the inner header is not decremented as the packet traverses the
enterprise. The VET interface therefore presents an automatic
tunneling abstraction that represents the enterprise as a single
IP hop.

Templin Informational [Page 8]

RFC 5558 VET February 2010

VET interfaces in non-multicast environments are Non-Broadcast,
Multiple Access (NBMA); VET interfaces in multicast environments
are multicast capable.

VET host
any node (host or router) that configures a VET interface for host
operation only. Note that a single node may configure some of its
VET interfaces as host interfaces and others as router interfaces.

VET node
any node that configures and uses a VET interface.

Provider-Independent (PI) prefix
an IPv6 or IPv4 prefix (e.g., 2001:DB8::/48, 192.0.2/24, etc.)
that is either self-generated by an ER or delegated to an
enterprise by a registry.

Provider Aggregated (PA) prefix
an IPv6 or IPv4 prefix that is delegated to an enterprise by a
provider network.

Routing Locator (RLOC)
a non-link-local IPv4 or IPv6 address taken from a PI/PA prefix
that can appear in enterprise-interior and/or interdomain routing
tables. Global-scope RLOC prefixes are delegated to specific
enterprises and are routable within both the enterprise-interior
and interdomain routing regions. Enterprise-local-scope RLOC
prefixes (e.g., IPv6 Unique Local Addresses [RFC4193], IPv4
privacy addresses [RFC1918], etc.) are self-generated by
individual enterprises and routable only within the enterprise-
interior routing region.

ERs use RLOCs for operating the enterprise-interior routing
protocol and for next-hop determination in forwarding packets
addressed to other RLOCs. End systems use RLOCs as addresses for
communications between endpoints within the same enterprise. VET
interfaces treat RLOCs as *outer* IP addresses during IP-in-IP
encapsulation.

Endpoint Interface iDentifier (EID)
an IPv4 or IPv6 address taken from a PI/PA prefix that is routable
within an enterprise-edge or VET overlay network scope, and may
also appear in enterprise-interior and/or interdomain mapping
tables. EID prefixes are typically separate and distinct from any
RLOC prefix space.

Templin Informational [Page 9]

RFC 5558 VET February 2010

Edge network routers use EIDs for operating the enterprise-edge or
VET overlay network routing protocol and for next-hop
determination in forwarding packets addressed to other EIDs. End
systems use EIDs as addresses for communications between endpoints
either within the same enterprise or within different enterprises.
VET interfaces treat EIDs as *inner* IP addresses during IP-in-IP
encapsulation.

The following additional acronyms are used throughout the document:

CGA - Cryptographically Generated Address
DHCP(v4, v6) - Dynamic Host Configuration Protocol
FIB - Forwarding Information Base
ISATAP - Intra-Site Automatic Tunnel Addressing Protocol
NBMA - Non-Broadcast, Multiple Access
ND - Neighbor Discovery
PIO - Prefix Information Option
PRL - Potential Router List
PRLNAME - Identifying name for the PRL (default is "isatap")
RIO - Route Information Option
RS/RA - IPv6 ND Router Solicitation/Advertisement
SEAL - Subnetwork Encapsulation and Adaptation Layer
SLAAC - IPv6 StateLess Address AutoConfiguation

3. Enterprise Characteristics

Enterprises consist of links that are connected by Enterprise Routers
(ERs) as depicted in Figure 1. ERs typically participate in a
routing protocol over enterprise-interior interfaces to discover
routes that may include multiple Layer 2 or Layer 3 forwarding hops.
Enterprise Border Routers (EBRs) are ERs that connect edge networks
to the enterprise and/or join multiple enterprises together.
Enterprise Border Gateways (EBGs) are EBRs that either directly or
indirectly connect enterprises to provider networks.

An enterprise may be as simple as a small collection of ERs and their
attached edge networks; an enterprise may also contain other
enterprises and/or be a subnetwork of a larger enterprise. An
enterprise may further encompass a set of branch offices and/or
nomadic hosts connected to a home office over one or several service
providers, e.g., through Virtual Private Network (VPN) tunnels.

Enterprises that comprise link types with sufficiently similar
properties (e.g., Layer 2 (L2) address formats, maximum transmission
units (MTUs), etc.) can configure a sub-IP layer routing service such
that IP sees the enterprise as an ordinary shared link the same as
for a (bridged) campus LAN. In that case, a single IP hop is
sufficient to traverse the enterprise without IP layer encapsulation.

Templin Informational [Page 10]

RFC 5558 VET February 2010

Enterprises that comprise link types with diverse properties and/or
configure multiple IP subnets must also provide a routing service
that operates as an IP layer mechanism. In that case, multiple IP
hops may be necessary to traverse the enterprise such that care must
be taken to avoid multi-link subnet issues [RFC4903].

Conceptually, an ER embodies both a host function and router
function. The host function supports Endpoint Interface iDentifier
(EID)-based and/or Routing LOCator (RLOC)-based communications
according to the weak end-system model [RFC1122]. The router
function engages in the enterprise-interior routing protocol,
connects any of the ER's edge networks to the enterprise, and may
also connect the enterprise to provider networks (see Figure 1).

In addition to other interface types, VET nodes configure VET
interfaces that view all other VET nodes in an enterprise as single-
hop neighbors attached to a virtual link. VET nodes configure a
separate VET interface for each distinct enterprise to which they
connect, and discover other EBRs on each VET interface that can be
used for forwarding packets to off-enterprise destinations.

For each distinct enterprise, an enterprise trust basis must be
established and consistently applied. For example, in enterprises in
which EBRs establish symmetric security associations, mechanisms such
as IPsec [RFC4301] can be used to assure authentication and
confidentiality. In other enterprise network scenarios, asymmetric
securing mechanisms such as SEcure Neighbor Discovery (SEND)
[RFC3971] may be necessary to authenticate exchanges based on trust
anchors.

Finally, in enterprises with a centralized management structure
(e.g., a corporate campus network), the enterprise name service and a
synchronized set of EBGs can provide infrastructure support for
virtual enterprise traversal. In that case, the EBGs can provide a
"default mapper" [APT] service used for short-term packet forwarding
until EBR neighbor relationships can be established. In enterprises
with a distributed management structure (e.g., MANETs), peer-to-peer
coordination between the EBRs themselves may be required.
Recognizing that various use cases will entail a continuum between a
fully distributed and fully centralized approach, the following
sections present the mechanisms of Virtual Enterprise Traversal as
they apply to a wide variety of scenarios.

4. Autoconfiguration

ERs, EBRs, EBGs, and VET hosts configure themselves for operation as
specified in the following subsections.

Templin Informational [Page 11]

RFC 5558 VET February 2010

4.1. Enterprise Router (ER) Autoconfiguration

ERs configure enterprise-interior interfaces and engage in any
routing protocols over those interfaces.

When an ER joins an enterprise, it first configures a unique IPv6
link-local address on each enterprise-interior interface and
configures an IPv4 link-local address on each enterprise-interior
interface that requires an IPv4 link-local capability. IPv6 link-
local address generation mechanisms that provide sufficient
uniqueness include Cryptographically Generated Addresses (CGAs)
[RFC3972], IPv6 Privacy Addresses [RFC4941], StateLess Address
AutoConfiguration (SLAAC) using EUI-64 interface identifiers
[RFC4291] [RFC4862], etc. The mechanisms specified in [RFC3927]
provide an IPv4 link-local address generation capability.

Next, the ER configures an RLOC on each of its enterprise-interior
interfaces and engages in any routing protocols on those interfaces.
The ER can configure an RLOC via explicit management, DHCP
autoconfiguration, pseudo-random self-generation from a suitably
large address pool, or through an alternate autoconfiguration
mechanism.

Alternatively (or in addition), the ER can request RLOC prefix
delegations via an automated prefix delegation exchange over an
enterprise-interior interface and can assign the prefix(es) on
enterprise-edge interfaces. In that case, the ER can use an RLOC
assigned to an enterprise-edge interface for enterprise-interior
routing protocol operation and next-hop determination purposes. Note
that in some cases, the same enterprise-edge interfaces may assign
both RLOC and an EID addresses if there is a means for source address
selection. In other cases (e.g., for separation of security
domains), RLOCs and EIDs must be assigned on separate sets of
enterprise-edge interfaces.

Self-generation of RLOCs for IPv6 can be from a large IPv6 local-use
address range, e.g., IPv6 Unique Local Addresses [RFC4193]. Self-
generation of RLOCs for IPv4 can be from a large IPv4 private address
range (e.g., [RFC1918]). When self-generation is used alone, the ER
must continuously monitor the RLOCs for uniqueness, e.g., by
monitoring the routing protocol.

DHCP generation of RLOCs may require support from relays within the
enterprise. For DHCPv6, relays that do not already know the RLOC of
a server within the enterprise forward requests to the
'All_DHCP_Servers' site-scoped IPv6 multicast group [RFC3315]. For
DHCPv4, relays that do not already know the RLOC of a server within
the enterprise forward requests to the site-scoped IPv4 multicast

Templin Informational [Page 12]

RFC 5558 VET February 2010

group address 'All_DHCPv4_Servers', which should be set to
239.255.2.1 unless an alternate multicast group for the site is
known. DHCPv4 servers that delegate RLOCs should therefore join the
'All_DHCPv4_Servers' multicast group and service any DHCPv4 messages
received for that group.

A combined approach using both DHCP and self-generation is also
possible when the ER configures both a DHCP client and relay that are
connected, e.g., via a pair of back-to-back connected Ethernet
interfaces, a tun/tap interface, a loopback interface, inter-process
communication, etc. The ER first self-generates a temporary RLOC
used only for the purpose of procuring an actual RLOC taken from a
disjoint addressing range. The ER then engages in the routing
protocol and performs a DHCP client/relay exchange using the
temporary RLOC as the address of the relay. When the DHCP server
delegates an actual RLOC address/prefix, the ER abandons the
temporary RLOC and re-engages in the routing protocol using an RLOC
taken from the delegation.

In some enterprise use cases (e.g., MANETs), assignment of RLOCs on
enterprise-interior interfaces as singleton addresses (i.e., as
addresses with /32 prefix lengths for IPv4, and as addresses with
/128 prefix lengths for IPv6) may be necessary to avoid multi-link
subnet issues.

4.2. Enterprise Border Router (EBR) Autoconfiguration

EBRs are ERs that configure VET interfaces over distinct sets of
underlying interfaces belonging to the same enterprise; an EBR can
connect to multiple enterprises, in which case it would configure
multiple VET interfaces. In addition to the ER autoconfiguration
procedures specified in Section 4.1, EBRs perform the following
autoconfiguration operations.

4.2.1. VET Interface Autoconfiguration

VET interface autoconfiguration entails:

1) interface initialization,
2) EBG discovery and enterprise identification, and
3) EID configuration.

These functions are specified in the following sections.

Templin Informational [Page 13]

RFC 5558 VET February 2010

4.2.1.1. Interface Initialization

EBRs configure a VET interface over a set of underlying interfaces
belonging to the same enterprise, where the VET interface presents a
virtual-link abstraction in which all EBRs in the enterprise appear
as single-hop neighbors through the use of IP-in-IP encapsulation.
After the EBR configures a VET interface, it initializes the
interface and assigns an IPv6 link-local address and an IPv4 link-
local address if necessary.

When IPv6 and IPv4 are used as the inner/outer protocols
(respectively), the EBR autoconfigures an ISATAP link-local address
([RFC5214], Section 6.2) on the VET interface to support packet
forwarding and operation of the IPv6 neighbor discovery protocol.
The ISATAP link-local address embeds an IPv4 RLOC, and need not be
checked for uniqueness since the IPv4 RLOC itself is managed for
uniqueness (see Section 4.1).

Link-local address configuration for other inner/outer IP protocol
combinations is through administrative configuration or through an
unspecified alternate method. Link-local address configuration for
other inner/outer IP protocol combinations may not be necessary if an
EID can be configured through other means (see Section 4.2.1.3).

After the EBR initializes a VET interface, it can communicate with
other VET nodes as single-hop neighbors on the VET interface from the
viewpoint of the inner IP protocol.

4.2.1.2. Enterprise Border Gateway Discovery and Enterprise
Identification

The EBR next discovers a list of EBGs for each of its VET interfaces.
The list can be discovered through information conveyed in the
routing protocol, through the Potential Router List (PRL) discovery
mechanisms outlined in Section 8.3.2 of [RFC5214], through DHCP
options, etc. In multicast-capable enterprises, EBRs can also listen
for advertisements on the 'rasadv' [RASADV] multicast group address.

In particular, whether or not routing information is available, the
EBR can discover the list of EBGs by resolving an identifying name
for the PRL ('PRLNAME') formed as 'hostname.domainname', where
'hostname' is an enterprise-specific name string and 'domainname' is
an enterprise-specific DNS suffix. The EBR discovers 'PRLNAME'
through manual configuration, a DHCP option, 'rasadv' protocol
advertisements, link-layer information (e.g., an IEEE 802.11 Service
Set Identifier (SSID)), or through some other means specific to the
enterprise. In the absence of other information, the EBR sets the

Templin Informational [Page 14]

RFC 5558 VET February 2010

'hostname' component of 'PRLNAME' to "isatap" and sets the
'domainname' component only if an enterprise-specific DNS suffix
"example.com" is known (e.g., as "isatap.example.com").

The global Internet interdomain routing core represents a specific
example of an enterprise network scenario, albeit on an enormous
scale. The 'PRLNAME' assigned to the global Internet interdomain
routing core is "isatap.net".

After discovering 'PRLNAME', the EBR can discover the list of EBGs by
resolving 'PRLNAME' to a list of RLOC addresses through a name
service lookup. For centrally managed enterprises, the EBR resolves
'PRLNAME' using an enterprise-local name service (e.g., the
enterprise-local DNS). For enterprises with a distributed management
structure, the EBR resolves 'PRLNAME' using Link-Local Multicast Name
Resolution (LLMNR) [RFC4795] over the VET interface. In that case,
all EBGs in the PRL respond to the LLMNR query, and the EBR accepts
the union of all responses.

Each distinct enterprise must have a unique identity that EBRs can
use to uniquely discern their enterprise affiliations. 'PRLNAME' as
well as the RLOCs of EBGs and the IP prefixes they aggregate serve as
an identifier for the enterprise.

4.2.1.3. EID Configuration

After EBG discovery, the EBR configures EIDs on its VET interfaces.
When IPv6 and IPv4 are used as the inner/outer protocols
(respectively), the EBR autoconfigures EIDs as specified in Section
5.4.1. In particular, the EBR acts as a host on its VET interfaces
for router and prefix discovery purposes but acts as a router on its
VET interfaces for routing protocol operation and packet forwarding
purposes.

EID configuration for other inner/outer IP protocol combinations is
through administrative configuration or through an unspecified
alternate method; in some cases, such EID configuration can be
performed independently of EBG discovery.

4.2.2. Provider-Aggregated (PA) EID Prefix Autoconfiguration

EBRs can acquire Provider-Aggregated (PA) EID prefixes through
autoconfiguration exchanges with EBGs over VET interfaces, where each
EBG may be configured as either a DHCP relay or DHCP server.

For IPv4 EIDs, the EBR acquires prefixes via an automated IPv4 prefix
delegation exchange, explicit management, etc.

Templin Informational [Page 15]

RFC 5558 VET February 2010

For IPv6 EIDs, the EBR acquires prefixes via DHCPv6 Prefix Delegation
exchanges. In particular, the EBR (acting as a requesting router)
can use DHCPv6 prefix delegation [RFC3633] over the VET interface to
obtain IPv6 EID prefixes from the server (acting as a delegating
router).

The EBR obtains prefixes using either a 2-message or 4-message DHCPv6
exchange [RFC3315]. For example, to perform the 2-message exchange,
the EBR's DHCPv6 client forwards a Solicit message with an IA_PD
option to its DHCPv6 relay, i.e., the EBR acts as a combined client/
relay (see Section 4.1). The relay then forwards the message over
the VET interface to an EBG, which either services the request or
relays it further. The forwarded Solicit message will elicit a reply
from the server containing PA IPv6 prefix delegations.

The EBR can propose a specific prefix to the DHCPv6 server per
Section 7 of [RFC3633], e.g., if a prefix delegation hint is
available. The server will check the proposed prefix for consistency
and uniqueness, then return it in the reply to the EBR if it was able
to perform the delegation.

After the EBR receives PA prefix delegations, it can provision the
prefixes on enterprise-edge interfaces as well as on other VET
interfaces for which it is configured as an EBG. It can also
provision the prefixes on enterprise-interior interfaces as long as
other nodes on those interfaces unambiguously associate the prefixes
with the EBR.

4.2.3. Provider-Independent (PI) EID Prefix Autoconfiguration

Independent of any PA prefixes, EBRs can acquire and use Provider-
Independent (PI) EID prefixes that are self-configured (e.g., using
[RFC4193], etc.) and/or delegated by a registration authority (e.g.,
using [CENTRL-ULA], etc.). When an EBR acquires a PI prefix, it must
also obtain credentials that it can use to prove prefix ownership
when it registers the prefixes with EBGs within an enterprise (see
Sections 5.4 and 5.5).

After the EBR receives PI prefix delegations, it can provision the
prefixes on enterprise-edge interfaces as well as on other VET
interfaces for which it is configured as an EBG. It can also
provision the prefixes on enterprise-interior interfaces as long as
other nodes on those interfaces can unambiguously associate the
prefixes with the EBR.

The minimum-sized IPv6 PI prefix that an EBR may acquire is a /56.

The minimum-sized IPv4 PI prefix that an EBR may acquire is a /24.

Templin Informational [Page 16]

RFC 5558 VET February 2010

4.3. Enterprise Border Gateway (EBG) Autoconfiguration

EBGs are EBRs that connect child enterprises to provider networks via
provider-edge interfaces and/or via VET interfaces configured over
parent enterprises. EBGs autoconfigure their provider-edge
interfaces in a manner that is specific to the provider connections,
and they autoconfigure their VET interfaces that were configured over
parent enterprises, using the EBR autoconfiguration procedures
specified in Section 4.2.

For each of its VET interfaces configured over a child enterprise,
the EBG initializes the interface and configures an EID the same as
for an ordinary EBR (see Section 4.2.1). It must then arrange to add
one or more of its RLOCs associated with the child enterprise to the
PRL, and it must maintain these resource records in accordance with
[RFC5214], Section 9. In particular, for each VET interface
configured over a child enterprise, the EBG adds the RLOCs to name-
service resource records for 'PRLNAME'.

EBGs respond to LLMNR queries for 'PRLNAME' on VET interfaces
configured over child enterprises with a distributed management
structure.

EBGs configure a DHCP relay/server on VET interfaces configured over
child enterprises that require DHCP services.

To avoid looping, EBGs must not configure a default route on a VET
interface configured over a child interface.

4.4. VET Host Autoconfiguration

Nodes that cannot be attached via an EBR's enterprise-edge interface
(e.g., nomadic laptops that connect to a home office via a Virtual
Private Network (VPN)) can instead be configured for operation as a
simple host connected to the VET interface. Such VET hosts perform
the same VET interface autoconfiguration procedures as specified for
EBRs in Section 4.2.1, but they configure their VET interfaces as
host interfaces (and not router interfaces). VET hosts can then send
packets to the EID addresses of other hosts on the VET interface, or
to off-enterprise EID destinations via a next-hop EBR.

Note that a node may be configured as a host on some VET interfaces
and as an EBR/EBG on other VET interfaces.

Templin Informational [Page 17]

RFC 5558 VET February 2010

5. Internetworking Operation

Following the autoconfiguration procedures specified in Section 4,
ERs, EBRs, EBGs, and VET hosts engage in normal internetworking
operations as discussed in the following sections.

5.1. Routing Protocol Participation

Following autoconfiguration, ERs engage in any RLOC-based IP routing
protocols and forward IP packets with RLOC addresses. EBRs can
additionally engage in any EID-based IP routing protocols and forward
IP packets with EID addresses. Note that the EID-based IP routing
domains are separate and distinct from any RLOC-based IP routing
domains.

5.2. RLOC-Based Communications

When permitted by policy and supported by routing, end systems can
avoid VET interface encapsulation through communications that
directly invoke the outer IP protocol using RLOC addresses instead of
EID addresses. End systems can use source address selection rules to
determine whether to use EID or RLOC addresses based on, e.g., name-
service records.

5.3. EID-Based Communications

In many enterprise scenarios, the use of EID-based communications
(i.e., instead of RLOC-based communications) may be necessary and/or
beneficial to support address scaling, NAT avoidance, security domain
separation, site multihoming, traffic engineering, etc.

The remainder of this section discusses internetworking operation for
EID-based communications using the VET interface abstraction.

5.4. IPv6 Router Discovery and Prefix Registration

The following sections discuss router and prefix discovery
considerations for the case of IPv6 as the inner IP protocol.

5.4.1. IPv6 Router and Prefix Discovery

EBGs follow the router and prefix discovery procedures specified in
[RFC5214], Section 8.2. They send solicited RAs over VET interfaces
for which they are configured as gateways with default router
lifetimes, with PIOs that contain PA prefixes for SLAAC, and with any
other required options/parameters. The RAs can also include PIOs
with the 'L' bit set to 0 and with a prefix such as '2001: DB8::/48'

Templin Informational [Page 18]

RFC 5558 VET February 2010

as a hint of an aggregated prefix from which the EBG is willing to
delegate longer PA prefixes. When PIOs that contain PA prefixes for
SLAAC are included, the 'M' flag in the RA should also be set to 0.

VET nodes follow the router and prefix discovery procedures specified
in [RFC5214], Section 8.3. They discover EBGs within the enterprise
as specified in Section 4.2.1.2, then perform RS/RA exchanges with
the EBGs to establish and maintain default routes. In particular,
the VET node sends unicast RS messages to EBGs over its VET
interface(s) to receive RAs. Depending on the enterprise network
trust basis, VET nodes may be required to use SEND to secure the
RS/RA exchanges.

When the VET node receives an RA, it authenticates the message, then
configures a default route based on the Router Lifetime. If the RA
contains Prefix Information Options (PIOs) with the 'A' and 'L' bits
set to 1, the VET node also autoconfigures IPv6 addresses from the
advertised prefixes using SLAAC and assigns them to the VET
interface. Thereafter, the VET node accepts packets that are
forwarded by EBGs for which it has current default routing
information (i.e., ingress filtering is based on the default router
trust relationship rather than a prefix-specific ingress filter
entry).

In enterprises in which DHCPv6 is preferred, DHCPv6 exchanges between
EBRs and EBGs may be sufficient to convey default router and prefix
information. In that case, RS/RA exchanges may not be necessary.

5.4.2. IPv6 PA Prefix Registration

After an EBR discovers default routes, it can use DHCP prefix
delegation to obtain PA prefixes via an EBG as specified in Section
4.2.2. The DHCP server ensures that the delegations are unique and
that the EBG's router function will forward IP packets over the VET
interface to the correct EBR. In particular, the EBG must register
and track the PA prefixes that are delegated to each EBR.

The PA prefix registrations remain active in the EBGs as long as the
EBR continues to issue DHCP renewals over the VET interface before
lease lifetimes expire. The lease lifetime also keeps the delegation
state active even if communications between the EBR and DHCP server
are disrupted for a period of time (e.g., due to an enterprise
network partition) before being reestablished (e.g., due to an
enterprise network merge).

Templin Informational [Page 19]

RFC 5558 VET February 2010

5.4.3. IPv6 PI Prefix Registration

After an EBR discovers default routes, it must register its PI
prefixes by sending RAs to a set of one or more EBGs with Route
Information Options (RIOs) [RFC4191] that contain the EBR's PI
prefixes. Each RA must include the RLOC of an EBG as the outer IP
destination address and a link-local address assigned to the VET
interface as the inner IP destination address. For enterprises that
use SEND, the RAs also include a CGA link-local inner source address,
SEND credentials, plus any certificates needed to prove ownership of
the PI prefixes. The EBR additionally tracks the set of EBGs to
which it sends RAs so that it can send subsequent RAs to the same
set.

When the EBG receives the RA, it first authenticates the message; if
the authentication fails, the EBG discards the RA. Otherwise, the
EBG installs the PI prefixes with their respective lifetimes in its
Forwarding Information Base (FIB) and configures them for both
ingress filtering [RFC3704] and forwarding purposes. In particular,
the EBG configures the FIB entries as ingress filter rules to accept
packets received on the VET interface that have a source address
taken from the PI prefixes. It also configures the FIB entries to
forward packets received on other interfaces with a destination
address taken from the PI prefixes to the EBR that registered the
prefixes on the VET interface.

The EBG then publishes the PI prefixes in a distributed database
(e.g., in a private instance of a routing protocol in which only EBGs
participate, via an automated name-service update mechanism
[RFC3007], etc.). For enterprises that are managed under a
centralized administrative authority, the EBG also publishes the PI
prefixes in the enterprise-local name-service (e.g., the enterprise-
local DNS [RFC1035]).

In particular, the EBG publishes each /56 prefix taken from the PI
prefixes as a separate Fully Qualified Domain Name (FQDN) that
consists of a sequence of 14 nibbles in reverse order (i.e., the same
as in [RFC3596], Section 2.5) followed by the string 'ip6' followed
by the string 'PRLNAME'. For example, when 'PRLNAME' is
"isatap.example.com", the EBG publishes the prefix '2001:DB8::/56'
as:

'0.0.0.0.0.0.8.b.d.0.1.0.0.2.ip6.isatap.example.com'.

The EBG includes the outer RLOC source address of the RA (e.g., in a
DNS A resource record) in each prefix publication. For enterprises
that use SEND, the EBG also includes the inner IPv6 CGA source
address (e.g., in a DNS AAAA record) in each prefix publication. If

Templin Informational [Page 20]

RFC 5558 VET February 2010

the prefix was already installed in the distributed database, the EBG
instead adds the outer RLOC source address (e.g., in an additional
DNS A record) to the preexisting publication to support PI prefixes
that are multihomed. For enterprises that use SEND, this latter
provision requires all EBRs of a multihomed site that advertise the
same PI prefixes in RAs to use the same CGA and the same SEND
credentials.

After the EBG authenticates the RA and publishes the PI prefixes, it
next acts as a Neighbor Discovery proxy (NDProxy) [RFC4389] on the
VET interfaces configured over any of its parent enterprises, and it
relays a proxied RA to the EBGs on those interfaces. (For
enterprises that use SEND, the EBG additionally acts as a SEcure
Neighbor Discovery Proxy (SENDProxy) [SEND-PROXY].) EBGs in parent
enterprises that receive the proxied RAs in turn act as
NDProxys/SENDProxys to relay the RAs to EBGs on their parent
enterprises, etc. The RA proxying and PI prefix publication recurses
in this fashion and ends when an EBR attached to an interdomain
routing core is reached.

After the initial PI prefix registration, the EBR that owns the
prefix(es) must periodically send additional RAs to its set of EBGs
to refresh prefix lifetimes. Each such EBG tracks the set of EBGs in
parent enterprises to which it relays the proxied RAs, and should
relay subsequent RAs to the same set.

This procedure has a direct analogy in the Teredo method of
maintaining state in network middleboxes through the periodic
transmission of "bubbles" [RFC4380].

5.4.4. IPv6 Next-Hop EBR Discovery

VET nodes discover destination-specific next-hop EBRs within the
enterprise by querying the name service for the /56 IPv6 PI prefix
taken from a packet's destination address, by forwarding packets via
a default route to an EBG, or by some other inner-IP-to-outer-IP
address mapping mechanism. For example, for the IPv6 destination
address '2001:DB8:1:2::1' and 'PRLNAME' "isatap.example.com" the VET
node can lookup the domain name:

'0.0.1.0.0.0.8.b.d.0.1.0.0.2.ip6.isatap.example.com'.

If the name-service lookup succeeds, it will return RLOC addresses
(e.g., in DNS A records) that correspond to next-hop EBRs to which
the VET node can forward packets. (In enterprises that use SEND, it
will also return an IPv6 CGA address, e.g., in a DNS AAAA record.)

Templin Informational [Page 21]

RFC 5558 VET February 2010

Name-service lookups in enterprises with a centralized management
structure use an infrastructure-based service, e.g., an enterprise-
local DNS. Name-service lookups in enterprises with a distributed
management structure and/or that lack an infrastructure-based name-
service instead use LLMNR over the VET interface. When LLMNR is
used, the EBR that performs the lookup sends an LLMNR query (with the
/56 prefix taken from the IP destination address encoded in dotted-
nibble format as shown above) and accepts the union of all replies it
receives from other EBRs on the VET interface. When an EBR receives
an LLMNR query, it responds to the query IFF it aggregates an IP
prefix that covers the prefix in the query.

Alternatively, in enterprises with a stable and highly-available set
of EBGs, the VET node can simply forward an initial packet via a
default route to an EBG. The EBG will forward the packet to a next-
hop EBR on the VET interface and return an ICMPv6 Redirect [RFC4861]
(using SEND, if necessary). If the packet's source address is on-
link on the VET interface, the EBG returns an ordinary "router-to-
host" redirect with the source address of the packet as its
destination. If the packet's source address is not on-link, the EBG
instead returns a "router-to-router" redirect with the link-local
ISATAP address of the previous-hop EBR as its destination. When IPv4
is used as the outer IP protocol, the EBG also includes in the
redirect one or more IPv6 Link-Layer Address Options (LLAOs) that
contain the IPv4 RLOCs of potential next-hop EBRs arranged in order
from lowest to highest priority (i.e., the first LLAO contains the
lowest priority RLOC and the final LLAO option contains the highest
priority). These LLAOs are formatted using a modified version of the
form specified in Section 5 of [RFC2529], as shown in Figure 2 (the
LLAO format for IPv6 as the outer IP protocol is out of scope).

+-------+-------+-------+-------+-------+-------+-------+-------+
| Type |Length | TTL | IPv4 Address |
+-------+-------+-------+-------+-------+-------+-------+-------+

Figure 2: VET Link-Layer Address Option Format

For each such IPv6/IPv4 LLAO, the Type is set to 2 (for Target Link-
Layer Address Option), Length is set to 1, and IPv4 Address is set to
the IPv4 RLOC of the next-hop EBR. TTL is set to the time in seconds
that the recipient may cache the RLOC, where the value 65535
represents infinity and the value 0 suspends forwarding through this
RLOC.

When a VET host receives an ordinary "router-to-host" redirect, it
processes the redirect exactly as specified in [RFC4861], Section 8.
When an EBR receives a "router-to-router" redirect, it discovers the
RLOC addresses of potential next-hop EBRs by examining the LLAOs

Templin Informational [Page 22]

RFC 5558 VET February 2010

included in the redirect. The EBR then installs a FIB entry that
contains the /56 prefix of the destination address encoded in the
redirect and the list of RLOCs of potential next-hop EBRs. The EBR
then enables the FIB entry for forwarding to next-hop EBRs but DOES
NOT enable it for ingress filtering acceptance of packets from next-
hop EBRs (i.e., the forwarding determination is unidirectional).

In enterprises in which spoofing is possible, after discovering
potential next-hop EBRs (either through name-service lookup or ICMP
redirect) the EBR must send authenticating credentials before
forwarding packets via the next-hops. To do so, the EBR must send
RAs over the VET interface (using SEND, if necessary) to one or more
of the potential next-hop EBRs with an RLOC as the outer IP
destination address. The RAs must include a Route Information Option
(RIO) [RFC4191] that contains the /56 PI prefix of the original
packet's source address. After sending the RAs, the EBR can either
enable the new FIB entry for forwarding immediately or delay until it
receives an explicit acknowledgement that a next-hop EBR received the
RA (e.g., using the SEAL explicit acknowledgement mechanism -- see
Section 5.7).

When a next-hop EBR receives the RA, it authenticates the message
then it performs a name-service lookup on the prefix in the RIO if
further authenticating evidence is required. If the name service
returns resource records that are consistent with the inner and outer
IP addresses of the RA, the next-hop EBR then installs the prefix in
the RIO in its FIB and enables the FIB entry for ingress filtering
but DOES NOT enable it for forwarding purposes. After an EBR sends
initial RAs following a redirect, it should send periodic RAs to
refresh the next-hop EBR's ingress filter prefix lifetimes as long as
traffic is flowing.

EBRs retain the FIB entries created as a result of an ICMP redirect
until all RLOC TTLs expire, or until no hints of forward progress
through any of the associated RLOCs are received. In this way, RLOC
liveness detection exactly parallels IPv6 Neighbor Unreachability
Detection ([RFC4861], Section 3).

5.5. IPv4 Router Discovery and Prefix Registration

When IPv4 is used as the inner IP protocol, router discovery and
prefix registration exactly parallel the mechanisms specified for
IPv6 in Section 5.4. To support this, modifications to the ICMPv4
Router Advertisement [RFC1256] function to include SEND constructs
and modifications to the ICMPv4 Redirect [RFC0792] function to
support router-to-router redirects will be specified in a future

Templin Informational [Page 23]

RFC 5558 VET February 2010

document. Additionally, publications for IPv4 prefixes will be in
dotted-nibble format in the 'ip4.isatap.example.com' domain. For
example, the IPv4 prefix 192.0.2/24 would be represented as:

'2.0.0.0.0.c.ip4.isatap.example.com'

5.6. VET Encapsulation

VET nodes forward packets by consulting the FIB to determine a
specific EBR/EBG as the next-hop router on a VET interface. When
multiple next-hop routers are available, VET nodes can use default
router preferences, routing protocol information, traffic engineering
configurations, etc. to select the best exit router. When there is
no FIB information other than "default" available, VET nodes can
discover the next-hop EBR/EBG through the mechanisms specified in
Section 5.4 and Section 5.5.

VET interfaces encapsulate inner IP packets in any mid-layer headers
followed by an outer IP header according to the specific
encapsulation type (e.g., [RFC4301], [RFC5214], [RFC5320], etc.);
they next submit the encapsulated packet to the outer IP forwarding
engine for transmission on an underlying interface.

For forwarding to next-hop addresses over VET interfaces that use
IPv6-in-IPv4 encapsulation, VET nodes determine the outer destination
address (i.e., the IPv4 RLOC of the next-hop EBR) through static
extraction of the IPv4 address embedded in the next-hop ISATAP
address. For other IP-in-IP encapsulations, determination of the
outer destination address is through administrative configuration or
through an unspecified alternate method. When there are multiple
candidate destination RLOCs available, the VET node should only
select an RLOC for which there is current forwarding information in
the outer IP protocol FIB.

5.7. SEAL Encapsulation

VET nodes should use SEAL encapsulation [RFC5320] over VET interfaces
to accommodate path MTU diversity, to defeat source address spoofing,
and to monitor next-hop EBR reachability. SEAL encapsulation
maintains a unidirectional and monotonically incrementing per-packet
identification value known as the 'SEAL_ID'. When a VET node that
uses SEAL encapsulation sends a SEND-protected Router Advertisement
(RA) or Router Solicitation (RS) message to another VET node, both
nodes cache the new SEAL_ID as per-tunnel state used for maintaining
a window of unacknowledged SEAL_IDs.

Templin Informational [Page 24]

RFC 5558 VET February 2010

In terms of security, when a VET node receives an ICMP message, it
can confirm that the packet-in-error within the ICMP message
corresponds to one of its recently sent packets by examining the
SEAL_ID along with source and destination addresses, etc.
Additionally, a next-hop EBR can track the SEAL_ID in packets
received from EBRs for which there is an ingress filter entry and
discard packets that have SEAL_ID values outside of the current
window.

In terms of next-hop reachability, an EBR can set the SEAL
"Acknowledgement Requested" bit in messages to receive confirmation
that a next-hop EBR is reachable. Setting the "Acknowledgement
Requested" bit is also used as the method for maintaining the window
of outstanding SEAL_IDs.

5.8. Generating Errors

When an EBR receives an IPv6 packet over a VET interface and there is
no matching ingress filter entry, it drops the packet and returns an
ICMPv6 [RFC4443] "Destination Unreachable; Source address failed
ingress/egress policy" message to the previous-hop EBR subject to
rate limiting.

When an EBR receives an IPv6 packet over a VET interface, and there
is no longest-prefix-match FIB entry for the destination, it returns
an ICMPv6 "Destination Unreachable; No route to destination" message
to the previous hop EBR subject to rate limiting.

When an EBR receives an IPv6 packet over a VET interface and the
longest-prefix-match FIB entry for the destination is via a next-hop
configured over the same VET interface the packet arrived on, the EBR
forwards the packet, then (if the FIB prefix is longer than ::/0)
sends a router-to-router ICMPv6 Redirect message (using SEND, if
necessary) to the previous-hop EBR as specified in Section 5.4.4.

Generation of other ICMP messages [RFC0792] [RFC4443] is the same as
for any IP interface.

5.9. Processing Errors

When an EBR receives an ICMPv6 "Destination Unreachable; Source
address failed ingress/egress policy" message from a next-hop EBR,
and there is a longest-prefix-match FIB entry for the original
packet's destination that is more specific than ::/0, the EBR
discards the message and marks the FIB entry for the destination as
"forwarding suspended" for the RLOC taken from the source address of
the ICMPv6 message. The EBR should then allow subsequent packets to
flow through different RLOCs associated with the FIB entry until it

Templin Informational [Page 25]

RFC 5558 VET February 2010

forwards a new RA to the suspended RLOC. If the EBR receives
excessive ICMPv6 ingress/egress policy errors through multiple RLOCs
associated with the same FIB entry, it should delete the FIB entry
and allow subsequent packets to flow through an EBG if supported in
the specific enterprise scenario.

When a VET node receives an ICMPv6 "Destination Unreachable; No route
to destination" message from a next-hop EBR, it forwards the ICMPv6
message to the source of the original packet as normal. If the EBR
has longest-prefix-match FIB entry for the original packet's
destination that is more specific than ::/0, the EBR also deletes the
FIB entry.

When an EBR receives an authentic ICMPv6 Redirect, it processes the
packet as specified in Section 5.4.4.

When an EBG receives new mapping information for a specific
destination prefix, it can propagate the update to other EBRs/EBGs by
sending an ICMPv6 redirect message to the 'All Routers' link-local
multicast address with an LLAO with the TTL for the unreachable LLAO
set to zero, and with a NULL packet in error.

Additionally, a VET node may receive ICMP "Destination Unreachable;
net / host unreachable" messages from an ER indicating that the path
to a VET neighbor may be failing. The VET node should first check,
e.g., the SEAL_ID, IPsec sequence number, source address of the
original packet if available, etc. to obtain reasonable assurance
that the ICMP message is authentic, then should mark the longest-
prefix-match FIB entry for the destination as "forwarding suspended"
for the RLOC destination address of the ICMP packet-in-error. If the
VET node receives excessive ICMP unreachable errors through multiple
RLOCs associated with the same FIB entry, it should delete the FIB
entry and allow subsequent packets to flow through a different route.

5.10. Mobility and Multihoming Considerations

EBRs that travel between distinct enterprise networks must either
abandon their PA prefixes that are relative to the "old" enterprise
and obtain new ones relative to the "new" enterprise or somehow
coordinate with a "home" enterprise to retain ownership of the
prefixes. In the first instance, the EBR would be required to
coordinate a network renumbering event using the new PA prefixes
[RFC4192]. In the second instance, an ancillary mobility management
mechanism must be used.

EBRs can retain their PI prefixes as they travel between distinct
enterprise networks as long as they register the prefixes with new
EBGs and (preferably) withdraw the prefixes from old EBGs prior to

Templin Informational [Page 26]

RFC 5558 VET February 2010

departure. Prefix registration with new EBGs is coordinated exactly
as specified in Section 5.4.3; prefix withdrawal from old EBGs is
simply through re-announcing the PI prefixes with zero lifetimes.

Since EBRs can move about independently of one another, stale FIB
entry state may be left in VET nodes when a neighboring EBR departs.
Additionally, EBRs can lose state for various reasons, e.g., power
failure, machine reboot, etc. For this reason, EBRs are advised to
set relatively short PI prefix lifetimes in RIO options, and to send
additional RAs to refresh lifetimes before they expire. (EBRs should
place conservative limits on the RAs they send to reduce congestion,
however.)

EBRs may register their PI prefixes with multiple EBGs for
multihoming purposes. EBRs should only forward packets via EBGs with
which it has registered its PI prefixes, since other EBGs may drop
the packets and return ICMPv6 "Destination Unreachable; Source
address failed ingress/egress policy" messages.

EBRs can also act as delegating routers to sub-delegate portions of
their PI prefixes to requesting routers on their enterprise-edge
interfaces and on VET interfaces for which they are configured as
EBGs. In this sense, the sub-delegations of an EBR's PI prefixes
become the PA prefixes for downstream-dependent nodes. Downstream-
dependent nodes that travel with a mobile provider EBR can continue
to use addresses configured from PA prefixes; downstream-dependent
nodes that move away from their provider EBR must perform address/
prefix renumbering when they associate with a new provider.

The EBGs of a multihomed enterprise should participate in a private
inner IP routing protocol instance between themselves (possibly over
an alternate topology) to accommodate enterprise partitions/merges as
well as intra-enterprise mobility events. These peer EBGs should
accept packets from one another without respect to the destination
(i.e., ingress filtering is based on the peering relationship rather
than a prefix-specific ingress filter entry).

5.11. Multicast

In multicast-capable deployments, ERs provide an enterprise-wide
multicasting service (e.g., Simplified Multicast Forwarding (SMF)
[MANET-SMF], Protocol Independent Multicast (PIM) routing, Distance
Vector Multicast Routing Protocol (DVMRP) routing, etc.) over their
enterprise-interior interfaces such that outer IP multicast messages
of site-scope or greater scope will be propagated across the
enterprise. For such deployments, VET nodes can also provide an
inner IP multicast/broadcast capability over their VET interfaces
through mapping of the inner IP multicast address space to the outer

Templin Informational [Page 27]

RFC 5558 VET February 2010

IP multicast address space. In that case, operation of link-scoped
(or greater scoped) inner IP multicasting services (e.g., a link-
scoped neighbor discovery protocol) over the VET interface is
available, but link-scoped services should be used sparingly to
minimize enterprise-wide flooding.

VET nodes encapsulate inner IP multicast messages sent over the VET
interface in any mid-layer headers (e.g., IPsec, SEAL, etc.) plus an
outer IP header with a site-scoped outer IP multicast address as the
destination. For the case of IPv6 and IPv4 as the inner/outer
protocols (respectively), [RFC2529] provides mappings from the IPv6
multicast address space to a site-scoped IPv4 multicast address space
(for other IP-in-IP encapsulations, mappings are established through
administrative configuration or through an unspecified alternate
static mapping).

Multicast mapping for inner IP multicast groups over outer IP
multicast groups can be accommodated, e.g., through VET interface
snooping of inner multicast group membership and routing protocol
control messages. To support inner-to-outer IP multicast mapping,
the VET interface acts as a virtual outer IP multicast host connected
to its underlying interfaces. When the VET interface detects that an
inner IP multicast group joins or leaves, it forwards corresponding
outer IP multicast group membership reports on an underlying
interface over which the VET interface is configured. If the VET
node is configured as an outer IP multicast router on the underlying
interfaces, the VET interface forwards locally looped-back group
membership reports to the outer IP multicast routing process. If the
VET node is configured as a simple outer IP multicast host, the VET
interface instead forwards actual group membership reports (e.g.,
IGMP messages) directly over an underlying interface.

Since inner IP multicast groups are mapped to site-scoped outer IP
multicast groups, the VET node must ensure that the site-scope outer
IP multicast messages received on the underlying interfaces for one
VET interface do not "leak out" to the underlying interfaces of
another VET interface. This is accommodated through normal site-
scoped outer IP multicast group filtering at enterprise boundaries.

5.12. Service Discovery

VET nodes can perform enterprise-wide service discovery using a
suitable name-to-address resolution service. Examples of flooding-
based services include the use of LLMNR [RFC4795] over the VET

Templin Informational [Page 28]

RFC 5558 VET February 2010

interface or multicast DNS [mDNS] over an underlying interface. More
scalable and efficient service discovery mechanisms are for further
study.

5.13. Enterprise Partitioning

EBGs can physically partition an enterprise by configuring multiple
VET interfaces over multiple distinct sets of underlying interfaces.
In that case, each partition (i.e., each VET interface) must
configure its own distinct 'PRLNAME' (e.g.,
'isatap.zone1.example.com', 'isatap.zone2.example.com', etc.).

EBGs can logically partition an enterprise using a single VET
interface by sending RAs with PIOs containing different IPv6 PA
prefixes to group nodes into different logical partitions. EBGs can
identify partitions, e.g., by examining RLOC prefixes, observing the
interfaces over which RSs are received, etc. In that case, a single
'PRLNAME' can cover all partitions.

5.14. EBG Prefix State Recovery

EBGs must retain explicit state that tracks the inner IP prefixes
owned by EBRs within the enterprise, e.g., so that packets are
delivered to the correct EBRs and not incorrectly "leaked out" of the
enterprise via a default route. For PA prefixes, the state is
maintained via an EBR's DHCP prefix delegation lease renewals, while
for PI prefixes the state is maintained via an EBR's periodic prefix
registration RAs.

When an EBG loses some or all of its state (e.g., due to a power
failure), it must recover the state so that packets can be forwarded
over correct routes. If the EBG aggregates PA prefixes from which
the IP prefixes of all EBRs in the enterprise are sub-delegated, then
the EBG can recover state through DHCP prefix delegation lease
renewals, through bulk lease queries, or through on-demand name-
service lookups based due to IP packet forwarding. If the EBG serves
as an anchor for PI prefixes, however, care must be taken to avoid
looping while state is recovered through prefix registration RAs from
EBRs. In that case, when the EBG that is recovering state forwards
an IP packet for which it has no explicit route other than ::/0, it
must first perform an on-demand name-service lookup to refresh state.

Templin Informational [Page 29]

RFC 5558 VET February 2010

6. Security Considerations

Security considerations for MANETs are found in [RFC2501].

Security considerations with tunneling that apply also to VET are
found in [RFC2529] [RFC5214]. In particular, VET nodes must verify
that the outer IP source address of a packet received on a VET
interface is correct for the inner IP source address using the
procedures specified in Section 7.3 of [RFC5214] in conjunction with
the ingress filtering mechanisms specified in this document.

SEND [RFC3971], IPsec [RFC4301], and SEAL [RFC5320] provide
additional securing mitigations to detect source address spoofing and
bogus RA messages sent by rogue routers.

Rogue routers can send bogus RA messages with spoofed RLOC source
addresses that can consume network resources and cause EBGs to
perform extra work. Nonetheless, EBGs should not "blacklist" such
RLOCs, as that may result in a denial of service to the RLOCs'
legitimate owners.

7. Related Work

Brian Carpenter and Cyndi Jung introduced the concept of intra-site
automatic tunneling in [RFC2529]; this concept was later called:
"Virtual Ethernet" and investigated by Quang Nguyen under the
guidance of Dr. Lixia Zhang. Subsequent works by these authors and
their colleagues have motivated a number of foundational concepts on
which this work is based.

Telcordia has proposed DHCP-related solutions for MANETs through the
CECOM MOSAIC program.

The Naval Research Lab (NRL) Information Technology Division uses
DHCP in their MANET research testbeds.

Security concerns pertaining to tunneling mechanisms are discussed in
[TUNNEL-SEC].

Default router and prefix information options for DHCPv6 are
discussed in [DEF-ROUTER].

An automated IPv4 prefix delegation mechanism is proposed in
[SUBNET].

RLOC prefix delegation for enterprise-edge interfaces is discussed in
[MANET-REC].

Templin Informational [Page 30]

RFC 5558 VET February 2010

MANET link types are discussed in [LINKTYPE].

Various proposals within the IETF have suggested similar mechanisms.

8. Acknowledgements

The following individuals gave direct and/or indirect input that was
essential to the work: Jari Arkko, Teco Boot, Emmanuel Bacelli, James
Bound, Scott Brim, Brian Carpenter, Thomas Clausen, Claudiu Danilov,
Ralph Droms, Dino Farinacci, Vince Fuller, Thomas Goff, Joel Halpern,
Bob Hinden, Sapumal Jayatissa, Dan Jen, Darrel Lewis, Tony Li, Joe
Macker, David Meyer, Thomas Narten, Pekka Nikander, Dave Oran,
Alexandru Petrescu, John Spence, Jinmei Tatuya, Dave Thaler, Ole
Troan, Michaela Vanderveen, Lixia Zhang, and others in the IETF
AUTOCONF and MANET working groups. Many others have provided
guidance over the course of many years.

9. Contributors

The following individuals have contributed to this document:

Eric Fleischman ([email protected])
Thomas Henderson ([email protected])
Steven Russert ([email protected])
Seung Yi ([email protected])

Ian Chakeres ([email protected]) contributed to earlier versions
of the document.

Jim Bound's foundational work on enterprise networks provided
significant guidance for this effort. We mourn his loss and honor
his contributions.

10. References

10.1. Normative References

[RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791,
September 1981.

[RFC0792] Postel, J., "Internet Control Message Protocol", STD 5,
RFC 792, September 1981.

[RFC0826] Plummer, D., "Ethernet Address Resolution Protocol: Or
Converting Network Protocol Addresses to 48.bit Ethernet
Address for Transmission on Ethernet Hardware", STD 37,
RFC 826, November 1982.

Templin Informational [Page 31]

RFC 5558 VET February 2010

[RFC1035] Mockapetris, P., "Domain names - implementation and
specification", STD 13, RFC 1035, November 1987.

[RFC2131] Droms, R., "Dynamic Host Configuration Protocol", RFC
2131, March 1997.

[RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6
(IPv6) Specification", RFC 2460, December 1998.

[RFC3007] Wellington, B., "Secure Domain Name System (DNS) Dynamic
Update", RFC 3007, November 2000.

[RFC3315] Droms, R., Ed., Bound, J., Volz, B., Lemon, T., Perkins,
C., and M. Carney, "Dynamic Host Configuration Protocol
for IPv6 (DHCPv6)", RFC 3315, July 2003.

[RFC3596] Thomson, S., Huitema, C., Ksinant, V., and M. Souissi,
"DNS Extensions to Support IP Version 6", RFC 3596,
October 2003.

[RFC3633] Troan, O. and R. Droms, "IPv6 Prefix Options for Dynamic
Host Configuration Protocol (DHCP) version 6", RFC 3633,
December 2003.

[RFC3971] Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander,
"SEcure Neighbor Discovery (SEND)", RFC 3971, March
2005.

[RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)",
RFC 3972, March 2005.

[RFC4191] Draves, R. and D. Thaler, "Default Router Preferences
and More-Specific Routes", RFC 4191, November 2005.

[RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing
Architecture", RFC 4291, February 2006.

[RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet
Control Message Protocol (ICMPv6) for the Internet
Protocol Version 6 (IPv6) Specification", RFC 4443,
March 2006.

[RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman,
"Neighbor Discovery for IP version 6 (IPv6)", RFC 4861,
September 2007.

[RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless
Address Autoconfiguration", RFC 4862, September 2007.

Templin Informational [Page 32]

RFC 5558 VET February 2010

[RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site
Automatic Tunnel Addressing Protocol (ISATAP)", RFC
5214, March 2008.

10.2. Informative References

[CATENET] Pouzin, L., "A Proposal for Interconnecting Packet
Switching Networks", May 1974.

[mDNS] Cheshire, S. and M. Krochmal, "Multicast DNS", Work in
Progress, September 2009.

[MANET-REC] Clausen, T. and U. Herberg, "MANET Router Configuration
Recommendations", Work in Progress, February 2009.

[LINKTYPE] Clausen, T., "The MANET Link Type", Work in Progress,
October 2008.

[DEF-ROUTER] Droms, R. and T. Narten, "Default Router and Prefix
Advertisement Options for DHCPv6", Work in Progress,
October 2009.

[SEND-PROXY] Krishnan, S., Laganier, J., and M. Bonola, "Secure Proxy
ND Support for SEND", Work in progress, July 2009.

[SUBNET] Johnson, R., Kumarasamy, J., Kinnear, K., and M. Stapp,
"Subnet Allocation Option", Work in Progress, October
2009.

[CENTRL-ULA] Hinden, R., Huston, G., and T. Narten, "Centrally
Assigned Unique Local IPv6 Unicast Addresses", Work in
Progress, June 2007.

[MANET-SMF] Macker, J., Ed. and SMF Design Team, "Simplified
Multicast Forwarding for MANET", Work in Progress, July
2009.

[TUNNEL-SEC] Hoagland, J., Krishnan, S., and D. Thaler, "Security
Concerns With IP Tunneling", Work in Progress, October
2008.

[APT] Jen, D., Meisel, M., Massey, D., Wang, L., Zhang, B.,
and L. Zhang, "APT: A Practical Transit Mapping
Service", Work in Progress, November 2007.

[IEN48] Cerf, V., "The Catenet Model for Internetworking", IEN
48, July 1978.

Templin Informational [Page 33]

RFC 5558 VET February 2010

[RASADV] Microsoft, "Remote Access Server Advertisement (RASADV)
Protocol Specification", October 2008.

[RFC1122] Braden, R., Ed., "Requirements for Internet Hosts -
Communication Layers", STD 3, RFC 1122, October 1989.

[RFC1256] Deering, S., Ed., "ICMP Router Discovery Messages", RFC
1256, September 1991.

[RFC1753] Chiappa, N., "IPng Technical Requirements Of the Nimrod
Routing and Addressing Architecture", RFC 1753, December
1994.

[RFC1918] Rekhter, Y., Moskowitz, B., Karrenberg, D., de Groot,
G., and E. Lear, "Address Allocation for Private
Internets", BCP 5, RFC 1918, February 1996.

[RFC1955] Hinden, R., "New Scheme for Internet Routing and
Addressing (ENCAPS) for IPNG", RFC 1955, June 1996.

[RFC2501] Corson, S. and J. Macker, "Mobile Ad hoc Networking
(MANET): Routing Protocol Performance Issues and
Evaluation Considerations", RFC 2501, January 1999.

[RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over
IPv4 Domains without Explicit Tunnels", RFC 2529, March
1999.

[RFC2775] Carpenter, B., "Internet Transparency", RFC 2775,
February 2000.

[RFC3704] Baker, F. and P. Savola, "Ingress Filtering for
Multihomed Networks", BCP 84, RFC 3704, March 2004.

[RFC3819] Karn, P., Ed., Bormann, C., Fairhurst, G., Grossman, D.,
Ludwig, R., Mahdavi, J., Montenegro, G., Touch, J., and
L. Wood, "Advice for Internet Subnetwork Designers", BCP
89, RFC 3819, July 2004.

[RFC3927] Cheshire, S., Aboba, B., and E. Guttman, "Dynamic
Configuration of IPv4 Link-Local Addresses", RFC 3927,
May 2005.

[RFC4192] Baker, F., Lear, E., and R. Droms, "Procedures for
Renumbering an IPv6 Network without a Flag Day", RFC
4192, September 2005.

Templin Informational [Page 34]

RFC 5558 VET February 2010

[RFC4193] Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast
Addresses", RFC 4193, October 2005.

[RFC4301] Kent, S. and K. Seo, "Security Architecture for the
Internet Protocol", RFC 4301, December 2005.

[RFC4380] Huitema, C., "Teredo: Tunneling IPv6 over UDP through
Network Address Translations (NATs)", RFC 4380, February
2006.

[RFC4389] Thaler, D., Talwar, M., and C. Patel, "Neighbor
Discovery Proxies (ND Proxy)", RFC 4389, April 2006.

[RFC4795] Aboba, B., Thaler, D., and L. Esibov, "Link-Local
Multicast Name Resolution (LLMNR)", RFC 4795, January
2007.

[RFC4852] Bound, J., Pouffary, Y., Klynsma, S., Chown, T., and D.
Green, "IPv6 Enterprise Network Analysis - IP Layer 3
Focus", RFC 4852, April 2007.

[RFC4903] Thaler, D., "Multi-Link Subnet Issues", RFC 4903, June
2007.

[RFC4941] Narten, T., Draves, R., and S. Krishnan, "Privacy
Extensions for Stateless Address Autoconfiguration in
IPv6", RFC 4941, September 2007.

[RFC5320] Templin, F., "The Subnetwork Encapsulation and
Adaptation Layer (SEAL)", RFC 5320, February 2010.

[RFC5720] Templin, F., "Routing and Addressing in Networks with
Global Enterprise Recursion (RANGER)", RFC 5720,
February 2010.

[RANGERS] Russert, S., Ed., Fleischman, E., Ed., and F. Templin,
Ed., "RANGER Scenarios", Work in Progress, September
2009.

Templin Informational [Page 35]

RFC 5558 VET February 2010

Appendix A. Duplicate Address Detection (DAD) Considerations

A priori uniqueness determination (also known as "pre-service DAD")
for an RLOC assigned on an enterprise-interior interface would
require either flooding the entire enterprise or somehow discovering
a link in the enterprise on which a node that configures a duplicate
address is attached and performing a localized DAD exchange on that
link. But, the control message overhead for such an enterprise-wide
DAD would be substantial and prone to false-negatives due to packet
loss and intermittent connectivity. An alternative to pre-service
DAD is to autoconfigure pseudo-random RLOCs on enterprise-interior
interfaces and employ a passive in-service DAD (e.g., one that
monitors routing protocol messages for duplicate assignments).

Pseudo-random IPv6 RLOCs can be generated with mechanisms such as
CGAs, IPv6 privacy addresses, etc. with very small probability of
collision. Pseudo-random IPv4 RLOCs can be generated through random
assignment from a suitably large IPv4 prefix space.

Consistent operational practices can assure uniqueness for EBG-
aggregated addresses/prefixes, while statistical properties for
pseudo-random address self-generation can assure uniqueness for the
RLOCs assigned on an ER's enterprise-interior interfaces. Still, an
RLOC delegation authority should be used when available, while a
passive in-service DAD mechanism should be used to detect RLOC
duplications when there is no RLOC delegation authority.

Author's Address

Fred L. Templin (editor)
Boeing Research & Technology
P.O. Box 3707 MC 7L-49
Seattle, WA 98124
USA

EMail: [email protected]

Templin Informational [Page 36]