| Network Working Group | F. Templin, Ed. |
| Internet-Draft | Boeing Research & Technology |
| Intended status: Informational | September 29, 2017 |
| Expires: April 2, 2018 |
IPv6 Prefix Delegation for Hosts
draft-templin-v6ops-pdhost-12.txt
IPv6 prefixes are typically delegated to requesting routers which then use them to number their downstream-attached links and networks. This document considers the case when the requesting router is a node that acts as a host on behalf of its local applications and as a router on behalf of any downstream networks.
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."
This Internet-Draft will expire on April 2, 2018.
Copyright (c) 2017 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 (https://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. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.
IPv6 Prefix Delegation (PD) entails 1) the communication of a prefix from a delegating router to a requesting router, 2) a representation of the prefix in the delegating router's routing table, and 3) a control messaging service between the delegating and requesting routers to maintain prefix lifetimes. Following delegation, the prefix is available for the requesting router's exclusive use and is not shared with any other nodes. An example IPv6 PD service is the Dynamic Host Configuration Protocol for IPv6 (DHCPv6) [RFC3315][RFC3633]. An alternative prefix management service based solely on IPv6 Neighbor Discovery (ND) messaging has also been proposed [I-D.pioxfolks-6man-pio-exclusive-bit].
This document considers the case when the requesting router is a node that acts as a host on behalf of its local applications and as a router on behalf of any downstream networks. The following paragraphs present possibilities for node behavior upon receipt of a delegated prefix.
For nodes that connect downstream-attached (aka "tethered") networks, a Delegating Router 'D' delegates a prefix 'P' to a Requesting node 'R' as shown in Figure 1:
+---------------------+
|Delegating Router 'D'|
| (Delegate 'P') |
+----------+----------+
|
| Upstream link
|
+----------+----------+
| Upstream Interface |
+---------------------+
| |
| Requesting node 'R' |
| (Receive 'P') |
| |
+--+-+--+-+--+-----+--+
|A1| |A2| |A3| ... |An|
+--+-+--+-+--+-----+--+
| Downstream Interface|
+----------+----------+
|
| Downstream link
|
X----+-------------+--------+----+---------------+---X
| | | |
+---++-+--+ +---++-+--+ +---++-+--+ +---++-+--+
| |X1| | | |X2| | | |X3| | | |Xn| |
| +--+ | | +--+ | | +--+ | | +--+ |
| Host H1 | | Host H2 | | Host H3 | ... | Host Hn |
+---------+ +---------+ +---------+ +---------+
<-------------- Tethered Network ------------->
Figure 1: Classic Routing Model
This document also considers the case when 'R' does not have any downstream interfaces, and can use 'P' solely for its own internal addressing purposes. In that case, 'R' assigns 'P' to a virtual interface (e.g., a loopback) that fills the role of a downstream interface.
'R' can then function under the weak end system (aka "weak host") model [RFC1122][RFC8028] by assigning addresses taken from 'P' to a virtual interface as shown in Figure 2:
+---------------------+
|Delegating Router 'D'|
| (Delegate 'P') |
+----------+----------+
|
| Upstream link
|
+----------+----------+
| Upstream Interface |
+---------------------+
| |
| Requesting node 'R' |
| (Receive 'P') |
| |
+--+-+--+-+--+-----+--+
|A1| |A2| |A3| ... |An|
+--+-+--+-+--+-----+--+
| Virtual Interface |
+---------------------+
Figure 2: Weak End System Model
'R' could instead function under the strong end system (aka "strong host") model [RFC1122][RFC8028] by assigning IPv6 addresses taken from 'P' to an upstream interface as shown in Figure 3:
+---------------------+
|Delegating Router 'D'|
| (Delegate 'P') |
+----------+----------+
|
| Upstream link
|
+----------+----------+
| Upstream Interface |
+--+-+--+-+--+-----+--+
|A1| |A2| |A3| ... |An|
+--+-+--+-+--+-----+--+
| |
| Requesting node 'R' |
| (Receive 'P') |
| |
+---------------------+
| Virtual Interface |
+---------------------+
Figure 3: Strong End System Model
The following sections present considerations for nodes that employ prefix delegation mechanisms.
The terminology of the normative references apply, and the terms "node", "host" and "router" are the same as defined in [RFC8200].
The following terms are defined for the purposes of this document:
IPv6 allows nodes to assign multiple addresses to a single interface. [RFC7934] discusses options for multi-addressing as well as use cases where multi-addressing may be desirable. Address configuration options for multi-addressing include StateLess Address AutoConfiguration (SLAAC) [RFC4862], DHCPv6 address configuration [RFC3315], manual configuration, etc.
Nodes configure addresses from a shared or individual prefix and assign them to the upstream interface over which the prefix was received. When the node assigns the addresses, it is required to use Multicast Listener Discovery (MLD) [RFC3810] to join the appropriate solicited-node multicast group(s) and to use the Duplicate Address Detection (DAD) algorithm [RFC4862] to ensure that no other node configures a duplicate address.
In contrast, a node that configures addresses from a delegated prefix can assign them without invoking MLD/DAD on an upstream interface, since the prefix has been delegated to the node for its own exclusive use and is not shared with any other nodes.
When a node receives a prefix delegation, it has many alternatives for provisioning the prefix. [RFC7278] discusses alternatives for provisioning a prefix obtained by a User Equipment (UE) device under the 3rd Generation Partnership Program (3GPP) service model. This document considers the more general case when the node receives a delegated prefix explicitly provided for its own exclusive use.
When the node receives the prefix, it can distribute the prefix to downstream networks and configure one or more addresses for itself on downstream interfaces. The node then acts as a router on behalf of its downstream networks and configures a default route via a neighbor on an upstream interface.
The node could instead (or in addition) use portions of the delegated prefix for its own multi-addressing purposes. In a first alternative, the node can assign as many addresses as it wants from the prefix to virtual interfaces. In that case, applications running on the node can use the addresses according to the weak end system model.
In a second alternative, the node can assign as many addresses as it wants from the prefix to the upstream interface over which the prefix was received. In that case, applications running on the node can use the addresses according to the strong end system model.
In both of these latter two cases, the node assigns the prefix itself to a virtual interface so that unused addresses from the prefix are correctly identified as unreachable. The node then acts as a host on behalf of its local applications even though neighbors on the upstream link see it as a router.
When a node configures addresses for itself from a shared or individual prefix, it performs MLD/DAD by sending multicast messages over upstream interfaces to test whether there is another node on the link that configures a duplicate address. When there are many such addresses and/or many such nodes, this could result in substantial multicast traffic that affects all nodes on the link.
When a node configures addresses for itself from a delegated prefix, it can configure as many addresses as it wants but does not perform MLD/DAD for any of the addresses over upstream interfaces. This means that the node can configure arbitrarily many addresses without causing any multicast messaging over the upstream interface that could disturb other nodes.
The node can be configured to either participate or not participate in a dynamic routing protocol over the upstream interface, according to the deployment model. When there are many nodes on the upstream link, dynamic routing protocol participation might be impractical due to scaling limitations, and may also be exacerbated by factors such as node mobility.
Unless it participates in a dynamic routing protocol, the node initially has only a default route pointing to a neighbor via an upstream interface. This means that packets sent by the node over an upstream interface will initially go through a default router even if there is a better first-hop node on the link.
The node acts as a simple host to send Router Solicitation (RS) messages over upstream interfaces (i.e., the same as described in Section 4.2 of [RFC7084]) but also sets the "Router" flag to TRUE in its Neighbor Advertisement messages. The node considers the upstream interfaces as non-advertising interfaces [RFC4861], i.e., it does not send RA messages over the upstream interfaces.
The current first-hop router may send a Redirect message that updates the node's neighbor cache so that future packets can use a better first-hop node on the link. The Redirect can apply either to a singleton destination address, or to an entire destination prefix as described in [I-D.templin-6man-rio-redirect].
The Internet Control Message Protocol for IPv6 (ICMPv6) includes a set of control message types [RFC4443] including Destination Unreachable (DU).
According to [RFC4443], routers should return DU messages (subject to rate limiting) with code 0 ("No route to destination") when a packet arrives for which there is no matching entry in the routing table, and with code 3 ("Address unreachable") when the IPv6 destination address cannot be resolved.
According to [RFC4443], hosts should return DU messages (subject to rate limiting) with code 3 to internal applications when the IPv6 destination address cannot be resolved, and with code 4 ("Port unreachable") if the IPv6 destination address is one of its own addresses but the transport protocol has no listener.
Nodes that obtain and manage prefix delegations per this document observe the same procedures as described for both routers and hosts above.
This document introduces no IANA considerations.
Security considerations for IPv6 Neighbor Discovery [RFC4861] and any applicable prefix delegation mechanisms apply to this document.
Additionally, the node may receive unwanted IPv6 packets via an upstream interface that match a delegated prefix but do not match either a configured IPv6 address or a transport listener. In that case, the node drops the packets and observes the "Destination Unreachable - Address/Port unreachable" procedures discussed in Section 8.
The node may also receive IPv6 packets via an upstream interface that do not match any of the node's delegated prefixes. In that case, the node drops the packets and observes the "Destination Unreachable - No route to destination" procedures discussed in Section 8. Dropping the packets is necessary to avoid a reflection attack that would cause the node to forward packets received from an upstream interface via the same or a different upstream interface.
In all cases, the node must decide whether or not to send DUs according to the specific operational scenario. In trusted networks, the node should send DU messages to provide useful information to potential correspondents. In untrusted networks, the node may refrain from sending DU messages to avoid providing sensitive information to potential attackers.
This work was motivated by discussions on the v6ops list. Mark Smith pointed out the need to consider MLD as well as DAD for the assignment of addresses to interfaces. Ricardo Pelaez-Negro, Edwin Cordeiro, Fred Baker, Naveen Lakshman, Ole Troan, Bob Hinden, Brian Carpenter, Joel Halpern and Albert Manfredi provided useful comments that have greatly improved the document.
This work is aligned with the NASA Safe Autonomous Systems Operation (SASO) program under NASA contract number NNA16BD84C.
This work is aligned with the FAA as per the SE2025 contract number DTFAWA-15-D-00030.
This work is aligned with the Boeing Information Technology (BIT) MobileNet program and the Boeing Research & Technology (BR&T) enterprise autonomy program.