IPv4 routes with an IPv6 next hop in the Babel routing protocol
IRIF, University of ParisCase 701475205 Paris Cedex 13Francejch@irif.frThis document defines an extension to the Babel routing protocol that
allows annoncing routes to an IPv4 prefix with an IPv6 next-hop, which
makes it possible for IPv4 traffic to flow through interfaces that have
not been assigned an IPv4 address.The role of a routing protocol is to build a routing table, a data
structure that maps network prefixes in a given family (IPv4 or IPv6) to
next hops, pairs of an outgoing interface and a neighbour's network
address, for example:When a packet is routed according to a given routing table entry, the
forwarding plane typically uses a neighbour discovery protocol (the
Neighbour Discovery protocol (ND) in the case of
IPv6, the Address Resolution Protocol (ARP) in
the case of IPv4) to map the next-hop address to a link-layer address (a
"MAC address"), which is then used to construct the link-layer frames that
encapsulate forwarded packets.It is apparent from the description above that there is no fundamental
reason why the destination prefix and the next-hop address should be in
the same address family: there is nothing preventing an IPv6 packet from
being routed through a next hop with an IPv4 address (in which case the
next hop's MAC address will be obtained using ARP), or, conversely, an
IPv4 packet from being routed through a next hop with an IPv6 address.
(In fact, it is even possible to store link-layer addresses directly in
the next-hop entry of the routing table, which is commonly done in
networks using the OSI protocol suite).The case of routing IPv4 packets through an IPv6 next hop is
particularly interesting, since it makes it possible to build networks
that have no IPv4 addresses except at the edges and still provide IPv4
connectivity to edge hosts. In addition, since an IPv6 next hop can use
a link-local address that is autonomously configured, the use of such
routes enables a mode of operation where the network core has no
statically assigned IP addresses of either family, which significantly
reduces the amount of manual configuration required.We call a route towards an IPv4 prefix that uses an IPv6 next hop
a "v4-via-v6" route. This document describes an extension that allows the
Babel routing protocol to announce v4-via-v6
routes across interfaces that have no IPv4 addresses assigned.
describes procedures that ensure that all routers
can originate ICMPv4 packets, even if they have not been assigned any IPv4
addresses.The extension described in this document is inspired by a previously
defined extension to the BGP protocol .The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
"OPTIONAL" in this document are to be interpreted as described in BCP 14
when, and only when,
they appear in all capitals, as shown here.The Babel protocol fully supports dual-stack operation: all data that
represent a neighbour address or a network prefix are tagged by an Address
Encoding (AE), a small integer that identifies the address family (IPv4 or
IPv6) of the address of prefix, and describes how it is encoded. This
extension defines a new AE, called v4-via-v6, which has the same format
as the existing AE for IPv4 addresses. This new AE is only allowed in
TLVs that carry network prefixes: TLVs that carry a neighbour address use
one of the normal encodings for IPv6 addresses.A Babel node that needs to announce an IPv4 route over an interface
that has no assigned IPv4 address MAY make a v4-via-v6 announcement. In
order to do so, it first establishes an IPv6 next-hop address in the usual
manner (either by sending the Babel packet over IPv6, or by including
a Next Hop TLV containing an IPv6 address and using AE 2 or 3); it then
sends an Update, with AE equal to 4 (v4-via-v6) containing the IPv4 prefix
being announced.If the outgoing interface has been assigned an IPv4 address, then, in
the interest of maximising compatibility with existing routers, the sender
SHOULD prefer an ordinary IPv4 announcement; even in that case, however,
it MAY send a v4-via-v6 announcement. A node SHOULD NOT send both
ordinary IPv4 and v4-via-v6 announcements for the same prefix over a single
interface (if the update is sent to a multicast address) or to a single
neighbour (if sent to a unicast address), since doing that provides no
benefit while doubling the amount of routing traffic.Updates with infinite metric are retractions: they indicate that
a previously announced route is no longer available. Retractions do not
require a next hop, and there is therefore no difference between v4-via-v6
retractions and ordinary retractions. A node MAY send IPv4 retractions
only, or it MAY send v4-via-v6 retractions on interfaces that have not
been assigned an IPv4 address.Upon reception of an Update TLV with AE equal to 4 (v4-via-v6) and
finite metric, a Babel node computes the IPv6 next hop, as described in
Section 4.6.9 of . If no IPv6 next hop exists,
then the Update MUST be silently ignored. If an IPv6 next hop exists,
then the node MAY acquire the route being announced, as described in
Section 3.5.3 of ; the parameters of the route are
as follows:
the prefix, plen, router-id, seqno, metric MUST be computed as for an
IPv4 route, as described in Section 4.6.9 of ;the next hop MUST be computed as for an IPv6 route, as described in
Section 4.6.9 of : it is taken from the last
preceding Next Hop TLV with an AE field equal to 2 or 3; if no such
entry exists, and if the Update TLV has been sent in a Babel packet
carried over IPv6, then the next hop is the network-layer source address
of the packet.An Update TLV with a v4-via-v6 AE and metric equal to infinity is
a retraction: it announces that a previously available route is being
retracted. In that case, no next hop is necessary, and the retraction is
treated as described in Section 4.6.9 of .As usual, a node MAY ignore the update, e.g., due to filtering
(Appendix C of ). If a node cannot install
v4-via-v6 routes, e.g., due to hardware or software limitations, then
routes to an IPv4 prefix with an IPv6 next hop MUST NOT be selected, as
described in Section 3.5.3 of .Prefix and seqno requests are used to request an update for a given
prefix. Since they are not related to a specific next hop, there is no
semantic difference between IPv4 and v4-via-v6 requests. Therefore,
a node SHOULD NOT send requests of either kind with the AE field being set
to 4 (v4-via-v6); instead, it SHOULD request IPv4 updates by sending
requests with the AE field being set to 1 (IPv4).When receiving requests, AEs 1 (IPv4) and 4 (v4-via-v6) MUST be treated
in the same manner: the receiver processes the request as described in
Section 3.8 of . If an Update is sent, then it
MAY be sent with AE 1 or 4, as described in
above, irrespective of which AE was used in the request.When receiving a request with AE 0 (wildcard), the receiver SHOULD send
a full route dump, as described in Section 3.8.1.1 of . Any IPv4 routes contained in the route dump MAY use
either AE 1 (IPv4) or AE 4 (v4-via-v6), as described in
above.The only other TLVs defined by that carry an
AE field are Next Hop and TLV. Next Hop and IHU TLVs MUST NOT carry the
AE 4 (v4-via-v6).The Internet Control Message Protocol (ICMPv4, or simply ICMP) is a protocol related to IPv4 that is primarily used to
carry diagnostic and debugging information. ICMPv4 packets may be
originated by end hosts (e.g., the "destination unreachable, port
unreachable" ICMPv4 packet), but they may also be originated by
intermediate routers (e.g., most other kinds of "destination unreachable"
packets).Some protocols deployed in the Internet rely on ICMPv4 packets sent by
intermediate routers. Most notably, path MTU Discovery (PMTUd) is an algorithm executed by end hosts to discover the
maximum packet size that a route is able to carry. While there exist
variants of PMTUd that are purely end-to-end , the
variant most commonly deployed in the Internet has a hard dependency on
ICMPv4 packets originated by intermediate routers: if intermediate routers
are unable to send ICMPv4 packets, PMTUd may lead to persistent
blackholing of IPv4 traffic.Due to this kind of dependency, every Babel router that is able to
forward IPv4 traffic MUST be able originate ICMPv4 traffic. Since the
extension described in this document enables routers to forward IPv4
traffic received over an interface that has not been assigned an IPv4
address, a router implementing this extension MUST be able to originate
ICMPv4 packets even when the outgoing interface has not been assigned an
IPv4 address.There are various ways to meet this requirement, and choosing between
them is left to the implementation. For example, if a router has an
interface that has been assigned an IPv4 address, or if an IPv4 address
has been assigned to the router itself (to the "loopback interface"), then
that IPv4 address may be "borrowed" to serve as the source of originated
ICMPv4 packets. If no IPv4 address is available, a router may choose
a source address from a prefix known to be unused, for example from
a suitably chosen private address range . If no
more suitable address is available, then a router MAY use the IPv4 dummy
address 192.0.0.8 as the source address of the IMCPv4 packets that it
sends. Note however that using the same address on multiple routers may
hamper debugging and fault isolation, e.g., when using the "traceroute"
utility.This extension defines the v4-via-v6 AE, whose value is 4. This AE is
solely used to tag network prefixes, and MUST NOT be used to tag neighbour
addresses, e.g. in Next Hop or IHU TLVs.This extension defines no new TLVs or sub-TLVs.Network prefixes tagged with AE 4 (v4-via-v6) MUST be encoded and
decoded just like prefixes tagged with AE 1 (IPv4), as described in
Section 4.3.1 of .A new compression state for AE 4 (v4-via-v6) distinct from that of AE
1 (IPv4) is introduced, and MUST be used for address compression of
prefixes tagged with AE 4, as described in Section 4.6.9 of
The following TLVs MAY be tagged with AE 4 (v4-via-v6):
Update (Type = 8)Route Request (Type = 9)Seqno Request (Type = 10)As AE 4 (v4-via-v6) is suitable only for network prefixes, IHU
(Type = 5) and Next-Hop (Type = 7) TLVs MUST NOT be
tagged with AE 4. Such (incorrect) TLVs MUST be ignored upon reception.An Update (Type = 8) TLV with AE 4 is constructed as described in
Section 4.6.9 of for AE 1 (IPv4), with the
following specificities:
Prefix. The Prefix field is constructed according to
above.Next Hop. The next hop is determined as described in
above.When tagged with the AE 4, Route Request and Seqno Request updates
MUST be constructed and decoded as described in Section 4.6 of
, and the network prefixes contained within them
decoded as described in above.This protocol extension adds no new TLVs or sub-TLVs.This protocol extension uses a new AE. As discussed in Appendix D of
and specified in the same document, implementations
that do not understand the present extension will silently ignore the various
TLVs that use this new AE. As a result, incompatible versions will ignore
v4-via-v6 routes. They will also ignore requests with AE 4, which, as
stated in , are NOT RECOMMENDED.Using a new AE introduces a new compression state, used to parse the
network prefixes. As this compression state is separate from other AEs'
states, it will not interfere with the compression state of unextended
nodes.This extension reuses the next-hop state from AEs 2 and 3 (IPv6), but
makes no changes to the way in which it is updated, and therefore causes
no compatibility issues.As mentioned in , ordinary IPv4 announcements
are preferred to v4-via-v6 announcements when the outgoing interface has
an assigned IPv4 address; doing otherwise would prevent routers that do
not implement this extension from learning the route being announced.IANA has allocated value 4 in the "Babel Address Encodings" registry as
follows:AENameReference4v4-via-v6(this document)The extension defined in this document does not fundamentally change
the security properties of the Babel protocol. However, by allowing IPv4
routes to be propagated across routers that have not been assigned IPv4
addresses, it might invalidate the assumptions made by network
administrators, which could conceivably lead to security issues.For example, if an island of IPv4-only hosts is separated from the IPv4
Internet by routers that have not been assigned IPv4 addresses, a network
administrator might reasonably assume that the IPv4-only hosts are
unreachable from the IPv4 Internet. This assumption is broken if the
intermediary routers implement the extension described in this document,
which might expose the IPv4-only hosts to traffic from the IPv4 Internet.
If this is undesirable, the flow of IPv4 traffic must be restricted by the
use of suitable filtering rules (Appendix C of )
together with matching packet filters in the data plane.This protocol extension was originally designed, described and
implemented in collaboration with Theophile Bastian. Margaret Cullen
pointed out the issues with ICMP and helped coin the phrase "v4-via-v6".
The author is also indebted to Donald Eastlake, Toke Hoiland-Jorgensen,
and David Schinazi.Key words for use in RFCs to Indicate Requirement LevelsAmbiguity of Uppercase vs Lowercase in RFC 2119 Key WordsThe Babel Routing ProtocolInternet Control Message ProtocolAn Ethernet Address Resolution Protocol: Or Converting Network
Protocol Addresses to 48.bit Ethernet Address for Transmission on Ethernet
HardwareNeighbor Discovery for IP version 6 (IPv6)
Advertising IPv4 Network Layer Reachability Information with an IPv6 Next Hop
Path MTU discoveryPacketization Layer Path MTU DiscoveryAddress Allocation for Private Internets