Network Working Group F. Templin, Ed.
Internet-Draft Boeing Research & Technology
Obsoletes: rfc5320, rfc5558, rfc5720, November 2, 2020
rfc6179, rfc6706 (if
approved)
Intended status: Standards Track
Expires: May 6, 2021
Asymmetric Extended Route Optimization (AERO)
draft-templin-intarea-6706bis-70
Abstract
This document specifies the operation of IP over Overlay Multilink
Network (OMNI) interfaces using the Asymmetric Extended Route
Optimization (AERO) internetworking and mobility management service.
AERO/OMNI use an IPv6 link-local address format that supports
operation of the IPv6 Neighbor Discovery (ND) protocol and links ND
to IP forwarding. Prefix delegation/registration services are
employed for network admission and to manage the routing system.
Multilink operation, mobility management, quality of service (QoS)
signaling and route optimization are naturally supported through
dynamic neighbor cache updates. Standard IP multicasting services
are also supported. AERO is a widely-applicable mobile
internetworking service especially well-suited to aviation services,
intelligent transportation systems, mobile Virtual Private Networks
(VPNs) and many other applications.
Status of This Memo
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 May 6, 2021.
Templin Expires May 6, 2021 [Page 1]
Internet-Draft AERO November 2020
Copyright Notice
Copyright (c) 2020 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.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 4
2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 5
3. Asymmetric Extended Route Optimization (AERO) . . . . . . . . 10
3.1. AERO Node Types . . . . . . . . . . . . . . . . . . . . . 10
3.2. The AERO Service over OMNI Links . . . . . . . . . . . . 12
3.2.1. AERO/OMNI Reference Model . . . . . . . . . . . . . . 12
3.2.2. Link-Local Addresses (LLAs) and Domain Local
Addresses (DLAs) . . . . . . . . . . . . . . . . . . 14
3.2.3. AERO Routing System . . . . . . . . . . . . . . . . . 15
3.2.4. OMNI Link Encapsulation . . . . . . . . . . . . . . . 16
3.2.5. Segment Routing Topologies (SRTs) . . . . . . . . . . 20
3.2.6. Segment Routing For OMNI Link Selection . . . . . . . 21
3.2.7. Segment Routing Within the OMNI Link . . . . . . . . 22
3.3. OMNI Interface Characteristics . . . . . . . . . . . . . 22
3.4. OMNI Interface Initialization . . . . . . . . . . . . . . 24
3.4.1. AERO Server/Relay Behavior . . . . . . . . . . . . . 24
3.4.2. AERO Proxy Behavior . . . . . . . . . . . . . . . . . 25
3.4.3. AERO Client Behavior . . . . . . . . . . . . . . . . 25
3.4.4. AERO Bridge Behavior . . . . . . . . . . . . . . . . 25
3.5. OMNI Interface Neighbor Cache Maintenance . . . . . . . . 25
3.5.1. OMNI Neighbor Interface Attributes . . . . . . . . . 27
3.5.2. OMNI Neighbor Advertisement Message Flags . . . . . . 28
3.6. OMNI Interface Encapsulation and Re-encapsulation . . . . 28
3.7. OMNI Interface Decapsulation . . . . . . . . . . . . . . 30
3.8. OMNI Interface Data Origin Authentication . . . . . . . . 30
3.9. OMNI Adaptation Layer and OMNI Interface MTU . . . . . . 30
3.10. OMNI Interface Forwarding Algorithm . . . . . . . . . . . 31
3.10.1. Client Forwarding Algorithm . . . . . . . . . . . . 32
3.10.2. Proxy Forwarding Algorithm . . . . . . . . . . . . . 32
3.10.3. Server/Relay Forwarding Algorithm . . . . . . . . . 33
3.10.4. Bridge Forwarding Algorithm . . . . . . . . . . . . 34
Templin Expires May 6, 2021 [Page 2]
Internet-Draft AERO November 2020
3.11. OMNI Interface Error Handling . . . . . . . . . . . . . . 35
3.12. AERO Router Discovery, Prefix Delegation and
Autoconfiguration . . . . . . . . . . . . . . . . . . . . 37
3.12.1. AERO Service Model . . . . . . . . . . . . . . . . . 38
3.12.2. AERO Client Behavior . . . . . . . . . . . . . . . . 38
3.12.3. AERO Server Behavior . . . . . . . . . . . . . . . . 40
3.13. The AERO Proxy . . . . . . . . . . . . . . . . . . . . . 43
3.13.1. Combined Proxy/Servers . . . . . . . . . . . . . . . 45
3.13.2. Detecting and Responding to Server Failures . . . . 45
3.13.3. Point-to-Multipoint Server Coordination . . . . . . 46
3.14. AERO Route Optimization / Address Resolution . . . . . . 47
3.14.1. Route Optimization Initiation . . . . . . . . . . . 47
3.14.2. Relaying the NS . . . . . . . . . . . . . . . . . . 48
3.14.3. Processing the NS and Sending the NA . . . . . . . . 48
3.14.4. Relaying the NA . . . . . . . . . . . . . . . . . . 49
3.14.5. Processing the NA . . . . . . . . . . . . . . . . . 49
3.14.6. Route Optimization Maintenance . . . . . . . . . . . 50
3.15. Neighbor Unreachability Detection (NUD) . . . . . . . . . 51
3.16. Mobility Management and Quality of Service (QoS) . . . . 52
3.16.1. Mobility Update Messaging . . . . . . . . . . . . . 53
3.16.2. Announcing Link-Layer Address and/or QoS Preference
Changes . . . . . . . . . . . . . . . . . . . . . . 54
3.16.3. Bringing New Links Into Service . . . . . . . . . . 54
3.16.4. Deactivating Existing Links . . . . . . . . . . . . 54
3.16.5. Moving Between Servers . . . . . . . . . . . . . . . 55
3.17. Multicast . . . . . . . . . . . . . . . . . . . . . . . . 56
3.17.1. Source-Specific Multicast (SSM) . . . . . . . . . . 56
3.17.2. Any-Source Multicast (ASM) . . . . . . . . . . . . . 58
3.17.3. Bi-Directional PIM (BIDIR-PIM) . . . . . . . . . . . 59
3.18. Operation over Multiple OMNI Links . . . . . . . . . . . 59
3.19. DNS Considerations . . . . . . . . . . . . . . . . . . . 60
3.20. Transition Considerations . . . . . . . . . . . . . . . . 60
3.21. Detecting and Reacting to Server and Bridge Failures . . 61
3.22. AERO Clients on the Open Internet . . . . . . . . . . . . 61
3.22.1. Use of SEND and CGA . . . . . . . . . . . . . . . . 64
3.23. Time-Varying MNPs . . . . . . . . . . . . . . . . . . . . 65
4. Implementation Status . . . . . . . . . . . . . . . . . . . . 66
5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 66
6. Security Considerations . . . . . . . . . . . . . . . . . . . 66
7. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 68
8. References . . . . . . . . . . . . . . . . . . . . . . . . . 69
8.1. Normative References . . . . . . . . . . . . . . . . . . 70
8.2. Informative References . . . . . . . . . . . . . . . . . 71
Appendix A. Non-Normative Considerations . . . . . . . . . . . . 77
A.1. Implementation Strategies for Route Optimization . . . . 77
A.2. Implicit Mobility Management . . . . . . . . . . . . . . 77
A.3. Direct Underlying Interfaces . . . . . . . . . . . . . . 78
A.4. AERO Critical Infrastructure Considerations . . . . . . . 78
Templin Expires May 6, 2021 [Page 3]
Internet-Draft AERO November 2020
A.5. AERO Server Failure Implications . . . . . . . . . . . . 79
A.6. AERO Client / Server Architecture . . . . . . . . . . . . 80
Appendix B. Change Log . . . . . . . . . . . . . . . . . . . . . 82
Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 83
1. Introduction
Asymmetric Extended Route Optimization (AERO) fulfills the
requirements of Distributed Mobility Management (DMM) [RFC7333] and
route optimization [RFC5522] for aeronautical networking and other
network mobility use cases such as intelligent transportation
systems. AERO is an internetworking and mobility management service
based on the Overlay Multilink Network Interface (OMNI)
[I-D.templin-6man-omni-interface] Non-Broadcast, Multiple Access
(NBMA) virtual link model. The OMNI link is a virtual overlay
configured over one or more underlying Internetworks, and nodes on
the link can exchange IP packets via tunneling. The OMNI Adaptation
Layer (OAL) supports multilink operation for increased reliability,
bandwidth optimization and traffic path selection while accommodating
Maximum Transmission Unit (MTU) diversity.
The AERO service comprises Clients, Proxys, Servers and Relays that
are seen as OMNI link neighbors as well as Bridges that interconnect
OMNI link segments. Each node's OMNI interface uses an IPv6 link-
local address format that supports operation of the IPv6 Neighbor
Discovery (ND) protocol [RFC4861] and links ND to IP forwarding. A
node's OMNI interface can be configured over multiple underlying
interfaces, and may therefore appear as a single interface with
multiple link-layer addresses. Each link-layer address is subject to
change due to mobility and/or QoS fluctuations, and link-layer
address changes are signaled by ND messaging the same as for any IPv6
link.
AERO provides a cloud-based service where mobile nodes may use any
Server acting as a Mobility Anchor Point (MAP) and fixed nodes may
use any Relay on the link for efficient communications. Fixed nodes
forward packets destined to other AERO nodes to the nearest Relay,
which forwards them through the cloud. A mobile node's initial
packets are forwarded through the Server, while direct routing is
supported through asymmetric extended route optimization while data
packets are flowing. Both unicast and multicast communications are
supported, and mobile nodes may efficiently move between locations
while maintaining continuous communications with correspondents and
without changing their IP Address.
AERO Bridges are interconnected in a secured private BGP overlay
routing instance using encapsulation to provide a hybrid routing/
bridging service that joins the underlying Internetworks of multiple
Templin Expires May 6, 2021 [Page 4]
Internet-Draft AERO November 2020
disjoint administrative domains into a single unified OMNI link.
Each OMNI link instance is characterized by the set of Mobility
Service Prefixes (MSPs) common to all mobile nodes. The link extends
to the point where a Relay/Server is on the optimal route from any
correspondent node on the link, and provides a conduit between the
underlying Internetwork and the OMNI link. To the underlying
Internetwork, the Relay/Server is the source of a route to the MSP,
and hence uplink traffic to the mobile node is naturally routed to
the nearest Relay/Server.
AERO assumes the use of PIM Sparse Mode in support of multicast
communication. In support of Source Specific Multicast (SSM) when a
Mobile Node is the source, AERO route optimization ensures that a
shortest-path multicast tree is established with provisions for
mobility and multilink operation. In all other multicast scenarios
there are no AERO dependencies.
AERO was designed for aeronautical networking for both manned and
unmanned aircraft, where the aircraft is treated as a mobile node
that can connect an Internet of Things (IoT). AERO is also
applicable to a wide variety of other use cases. For example, it can
be used to coordinate the Virtual Private Network (VPN) links of
mobile nodes (e.g., cellphones, tablets, laptop computers, etc.) that
connect into a home enterprise network via public access networks
using services such as OpenVPN [OVPN]. It can also be used to
facilitate vehicular and pedestrian communications services for
intelligent transportation systems. Other applicable use cases are
also in scope.
The following numbered sections present the AERO specification. The
appendices at the end of the document are non-normative.
2. Terminology
The terminology in the normative references applies; especially, the
terminology in the OMNI specification
[I-D.templin-6man-omni-interface] is used extensively throughout.
The following terms are defined within the scope of this document:
IPv6 Neighbor Discovery (ND)
an IPv6 control message service for coordinating neighbor
relationships between nodes connected to a common link. AERO uses
the ND service specified in [RFC4861].
IPv6 Prefix Delegation
a networking service for delegating IPv6 prefixes to nodes on the
link. The nominal service is DHCPv6 [RFC8415], however alternate
services (e.g., based on ND messaging) are also in scope. Most
Templin Expires May 6, 2021 [Page 5]
Internet-Draft AERO November 2020
notably, a minimal form of prefix delegation known as "prefix
registration" can be used if the Client knows its prefix in
advance and can represent it in the IPv6 source address of an ND
message.
Access Network (ANET)
a node's first-hop data link service network (e.g., a radio access
network, cellular service provider network, corporate enterprise
network, etc.) that often provides link-layer security services
such as IEEE 802.1X and physical-layer security prevent
unauthorized access internally and with border network-layer
security services such as firewalls and proxies that prevent
unauthorized outside access.
ANET interface
a node's attachment to a link in an ANET.
Internetwork (INET)
a connected IP network topology with a coherent routing and
addressing plan and that provides a transit backbone service for
ANET end systems. INETs also provide an underlay service over
which the AERO virtual link is configured. Example INETs include
corporate enterprise networks, aviation networks, and the public
Internet itself. When there is no administrative boundary between
an ANET and the INET, the ANET and INET are one and the same.
INET Partition
frequently, INETs such as large corporate enterprise networks are
sub-divided internally into separate isolated partitions. Each
partition is fully connected internally but disconnected from
other partitions, and there is no requirement that separate
partitions maintain consistent Internet Protocol and/or addressing
plans. (Each INET partition is seen as a separate OMNI link
segment as discussed below.)
INET interface
a node's attachment to a link in an INET.
INET address
an IP address assigned to a node's interface connection to an
INET.
INET encapsulation
the encapsulation of a packet in an outer header or headers that
can be routed within the scope of the local INET partition.
OMNI link
Templin Expires May 6, 2021 [Page 6]
Internet-Draft AERO November 2020
the same as defined in [I-D.templin-6man-omni-interface], and
manifested by IPv6 encapsulation [RFC2473]. The OMNI link spans
underlying INET segments joined by virtual bridges in a spanning
tree the same as a bridged campus LAN. AERO nodes on the OMNI
link appear as single-hop neighbors even though they may be
separated by multiple underlying INET hops, and can use Segment
Routing [RFC8402] to cause packets to visit selected waypoints on
the link.
OMNI domain
a set of affiliated OMNI links that collectively provide services
under a common (set of) Mobility Service Prefixes (MSPs).
OMNI Interface
a node's attachment to an OMNI link. Since the addresses assigned
to an OMNI interface are managed for uniqueness, OMNI interfaces
do not require Duplicate Address Detection (DAD) and therefore set
the administrative variable 'DupAddrDetectTransmits' to zero
[RFC4862].
OMNI Adaptation Layer (OAL)
an OMNI interface process whereby packets admitted into the
interface are wrapped in a mid-layer IPv6 header and fragmented/
reassembled if necessary to support the OMNI link Maximum
Transmission Unit (MTU). The OAL is also responsible for
generating MTU-related control messages as necessary, and for
providing addressing context for spanning multiple segments of a
bridged OMNI link.
OMNI Link-Local Address (LLA)
a link local IPv6 address per [RFC4291] constructed as specified
in Section 3.2.2.
OMNI Domain-Local Address (DLA)
an IPv6 address from the prefix [DLA]::/10 constructed as
specified in [I-D.templin-6man-omni-interface]. OMNI DLAs are
statelessly derived from OMNI LLAs, and vice-versa.
underlying interface
an ANET or INET interface over which an OMNI interface is
configured.
Mobility Service Prefix (MSP)
an IP prefix assigned to the OMNI link and from which more-
specific Mobile Network Prefixes (MNPs) are derived.
Mobile Network Prefix (MNP)
Templin Expires May 6, 2021 [Page 7]
Internet-Draft AERO November 2020
an IP prefix allocated from an MSP and delegated to an AERO Client
or Relay.
AERO node
a node that is connected to an OMNI link and participates in the
AERO internetworking and mobility service.
AERO Client ("Client")
an AERO node that connects over one or more underlying interfaces
and requests MNP delegation/registration service from AERO
Servers. The Client assigns a Client LLA to the OMNI interface
for use in ND exchanges with other AERO nodes and forwards packets
to correspondents according to OMNI interface neighbor cache
state.
AERO Server ("Server")
an INET node that configures an OMNI interface to provide default
forwarding and mobility/multilink services for AERO Clients. The
Server assigns an administratively-provisioned LLA to its OMNI
interface to support the operation of the ND services, and
advertises all of its associated MNPs via BGP peerings with
Bridges.
AERO Relay ("Relay")
an AERO Server that also provides forwarding services between
nodes reached via the OMNI link and correspondents on other links.
AERO Relays are provisioned with MNPs (i.e., the same as for an
AERO Client) and run a dynamic routing protocol to discover any
non-MNP IP routes. In both cases, the Relay advertises the MSP(s)
to its downstream networks, and distributes all of its associated
MNPs and non-MNP IP routes via BGP peerings with Bridges (i.e.,
the same as for an AERO Server).
AERO Bridge ("Bridge")
a node that provides hybrid routing/bridging services (as well as
a security trust anchor) for nodes on an OMNI link. As a router,
the Bridge forwards packets using standard IP forwarding. As a
bridge, the Bridge forwards packets over the OMNI link without
decrementing the IPv6 Hop Limit. AERO Bridges peer with Servers
and other Bridges to discover the full set of MNPs for the link as
well as any non-MNPs that are reachable via Relays.
AERO Proxy ("Proxy")
a node that provides proxying services between Clients in an ANET
and Servers in external INETs. The AERO Proxy is a conduit
between the ANET and external INETs in the same manner as for
common web proxies, and behaves in a similar fashion as for ND
proxies [RFC4389]. A node may be configured to act as either a
Templin Expires May 6, 2021 [Page 8]
Internet-Draft AERO November 2020
Proxy and/or a Server, depending on Client Server selection
criteria.
ingress tunnel endpoint (ITE)
an OMNI interface endpoint that injects encapsulated packets into
an OMNI link.
egress tunnel endpoint (ETE)
an OMNI interface endpoint that receives encapsulated packets from
an OMNI link.
link-layer address
an IP address used as an encapsulation header source or
destination address from the perspective of the OMNI interface.
When an upper layer protocol (e.g., UDP) is used as part of the
encapsulation, the port number is also considered as part of the
link-layer address.
network layer address
the source or destination address of an encapsulated IP packet
presented to the OMNI interface.
end user network (EUN)
an internal virtual or external edge IP network that an AERO
Client or Relay connects to the rest of the network via the OMNI
interface. The Client/Relay sees each EUN as a "downstream"
network, and sees the OMNI interface as the point of attachment to
the "upstream" network.
Mobile Node (MN)
an AERO Client and all of its downstream-attached networks that
move together as a single unit, i.e., an end system that connects
an Internet of Things.
Mobile Router (MR)
a MN's on-board router that forwards packets between any
downstream-attached networks and the OMNI link.
Route Optimization Source (ROS)
the AERO node nearest the source that initiates route
optimization. The ROS may be a Server or Proxy acting on behalf
of the source Client.
Route Optimization responder (ROR)
the AERO node nearest the target destination that responds to
route optimization requests. The ROR may be a Server acting on
behalf of a target MNP Client, or a Relay for a non-MNP
destination.
Templin Expires May 6, 2021 [Page 9]
Internet-Draft AERO November 2020
MAP List
a geographically and/or topologically referenced list of addresses
of all Servers within the same OMNI link. There is a single MAP
list for the entire OMNI link.
Distributed Mobility Management (DMM)
a BGP-based overlay routing service coordinated by Servers and
Bridges that tracks all Server-to-Client associations.
Mobility Service (MS)
the collective set of all Servers, Proxys, Bridges and Relays that
provide the AERO Service to Clients.
Mobility Service Endpoint MSE)
an individual Server, Proxy, Bridge or Relay in the Mobility
Service.
Throughout the document, the simple terms "Client", "Server",
"Bridge", "Proxy" and "Relay" refer to "AERO Client", "AERO Server",
"AERO Bridge", "AERO Proxy" and "AERO Relay", respectively.
Capitalization is used to distinguish these terms from other common
Internetworking uses in which they appear without capitalization.
The terminology of DHCPv6 [RFC8415] and IPv6 ND [RFC4861] (including
the names of node variables, messages and protocol constants) is used
throughout this document. The terms "All-Routers multicast", "All-
Nodes multicast", "Solicited-Node multicast" and "Subnet-Router
anycast" are defined in [RFC4291]. Also, the term "IP" is used to
generically refer to either Internet Protocol version, i.e., IPv4
[RFC0791] or IPv6 [RFC8200].
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 [RFC2119][RFC8174] when, and only when, they appear in all
capitals, as shown here.
3. Asymmetric Extended Route Optimization (AERO)
The following sections specify the operation of IP over OMNI links
using the AERO service:
3.1. AERO Node Types
AERO Clients are Mobile Nodes (MNs) that connect via underlying
interfaces with addresses that may change when the Client moves to a
new network connection point. AERO Clients register their Mobile
Network Prefixes (MNPs) with the AERO service, and distribute the
Templin Expires May 6, 2021 [Page 10]
Internet-Draft AERO November 2020
MNPs to nodes on EUNs. AERO Bridges, Servers, Proxys and Relays are
critical infrastructure elements in fixed (i.e., non-mobile) INET
deployments and hence have permanent and unchanging INET addresses.
Together, they constitute the AERO service which provides an OMNI
link virtual overlay for connecting AERO Clients.
AERO Bridges provide hybrid routing/bridging services (as well as a
security trust anchor) for nodes on an OMNI link. Bridges use
standard IPv6 routing to forward packets both within the same INET
partitions and between disjoint INET partitions based on a mid-layer
IPv6 encapsulation per [RFC2473]. The inner IP layer experiences a
virtual bridging service since the inner IP TTL/Hop Limit is not
decremented during forwarding. Each Bridge also peers with Servers
and other Bridges in a dynamic routing protocol instance to provide a
Distributed Mobility Management (DMM) service for the list of active
MNPs (see Section 3.2.3). Bridges present the OMNI link as a set of
one or more Mobility Service Prefixes (MSPs) and configure secured
tunnels with Servers, Relays, Proxys and other Bridges; they further
maintain IP forwarding table entries for each MNP and any other
reachable non-MNP prefixes.
AERO Servers provide default forwarding and mobility/multilink
services for AERO Client Mobile Nodes (MNs). Each Server also peers
with Bridges in a dynamic routing protocol instance to advertise its
list of associated MNPs (see Section 3.2.3). Servers facilitate
prefix delegation/registration exchanges with Clients, where each
delegated prefix becomes an MNP taken from an MSP. Servers forward
packets between OMNI interface neighbors and track each Client's
mobility profiles. Servers may further act as Servers for some sets
of Clients and as Proxies for others.
AERO Proxys provide a conduit for ANET Clients to associate with
Servers in external INETs. Client and Servers exchange control plane
messages via the Proxy acting as a bridge between the ANET/INET
boundary. The Proxy forwards data packets between Clients and the
OMNI link according to forwarding information in the neighbor cache.
The Proxy function is specified in Section 3.13. Proxys may further
act as Proxys for some sets of Clients and as Servers for others.
AERO Relays are Servers that provide forwarding services between the
OMNI interface and INET/EUN interfaces. Relays are provisioned with
MNPs the same as for an AERO Client, and also run a dynamic routing
protocol to discover any non-MNP IP routes. The Relay advertises the
MSP(s) to its connected networks, and distributes all of its
associated MNPs and non-MNP IP routes via BGP peerings with Bridges
Templin Expires May 6, 2021 [Page 11]
Internet-Draft AERO November 2020
3.2. The AERO Service over OMNI Links
3.2.1. AERO/OMNI Reference Model
Figure 1 presents the basic OMNI link reference model:
+----------------+
| AERO Bridge B1 |
| Nbr: S1, S2, P1|
|(X1->S1; X2->S2)|
| MSP M1 |
+-+---------+--+-+
+--------------+ | Secured | | +--------------+
|AERO Server S1| | tunnels | | |AERO Server S2|
| Nbr: C1, B1 +-----+ | +-----+ Nbr: C2, B1 |
| default->B1 | | | default->B1 |
| X1->C1 | | | X2->C2 |
+-------+------+ | +------+-------+
| OMNI link | |
X===+===+===================+==)===============+===+===X
| | | |
+-----+--------+ +--------+--+-----+ +--------+-----+
|AERO Client C1| | AERO Proxy P1 | |AERO Client C2|
| Nbr: S1 | |(Proxy Nbr Cache)| | Nbr: S2 |
| default->S1 | +--------+--------+ | default->S2 |
| MNP X1 | | | MNP X2 |
+------+-------+ .--------+------. +-----+--------+
| (- Proxyed Clients -) |
.-. `---------------' .-.
,-( _)-. ,-( _)-.
.-(_ IP )-. +-------+ +-------+ .-(_ IP )-.
(__ EUN )--|Host H1| |Host H2|--(__ EUN )
`-(______)-' +-------+ +-------+ `-(______)-'
Figure 1: AERO/OMNI Reference Model
In this model:
o the OMNI link is an overlay network service configured over one or
more underlying INET partitions which may be managed by different
administrative authorities and have incompatible protocols and/or
addressing plans.
o AERO Bridge B1 aggregates Mobility Service Prefix (MSP) M1,
discovers Mobile Network Prefixes (MNPs) X* and advertises the MSP
via BGP peerings over secured tunnels to Servers (S1, S2).
Bridges connect the disjoint segments of a partitioned OMNI link.
Templin Expires May 6, 2021 [Page 12]
Internet-Draft AERO November 2020
o AERO Servers/Relays S1 and S2 configure secured tunnels with
Bridge B1 and also provide mobility, multilink and default router
services for their associated Clients C1 and C2.
o AERO Clients C1 and C2 associate with Servers S1 and S2,
respectively. They receive Mobile Network Prefix (MNP)
delegations X1 and X2, and also act as default routers for their
associated physical or internal virtual EUNs. Simple hosts H1 and
H2 attach to the EUNs served by Clients C1 and C2, respectively.
o AERO Proxy P1 configures a secured tunnel with Bridge B1 and
provides proxy services for AERO Clients in secured enclaves that
cannot associate directly with other OMNI link neighbors.
An OMNI link configured over a single INET appears as a single
unified link with a consistent underlying network addressing plan.
In that case, all nodes on the link can exchange packets via simple
INET encapsulation, since the underlying INET is connected. In
common practice, however, an OMNI link may be partitioned into
multiple "segments", where each segment is a distinct INET
potentially managed under a different administrative authority (e.g.,
as for worldwide aviation service providers such as ARINC, SITA,
Inmarsat, etc.). Individual INETs may also themselves be partitioned
internally, in which case each internal partition is seen as a
separate segment.
The addressing plan of each segment is consistent internally but will
often bear no relation to the addressing plans of other segments.
Each segment is also likely to be separated from others by network
security devices (e.g., firewalls, proxies, packet filtering
gateways, etc.), and in many cases disjoint segments may not even
have any common physical link connections. Therefore, nodes can only
be assured of exchanging packets directly with correspondents in the
same segment, and not with those in other segments. The only means
for joining the segments therefore is through inter-domain peerings
between AERO Bridges.
The same as for traditional campus LANs, multiple OMNI link segments
can be joined into a single unified link via a virtual bridging
service using the OMNI Adaptation Layer (OAL) which inserts a mid-
layer IPv6 encapsulation per [RFC2473] that supports inter-segment
forwarding (i.e., bridging) without decrementing the network-layer
TTL/Hop Limit. This bridging of OMNI link segments is shown in
Figure 2:
Templin Expires May 6, 2021 [Page 13]
Internet-Draft AERO November 2020
. . . . . . . . . . . . . . . . . . . . . . .
. .
. .-(::::::::) .
. .-(::::::::::::)-. +-+ .
. (:::: Segment A :::)--|B|---+ .
. `-(::::::::::::)-' +-+ | .
. `-(::::::)-' | .
. | .
. .-(::::::::) | .
. .-(::::::::::::)-. +-+ | .
. (:::: Segment B :::)--|B|---+ .
. `-(::::::::::::)-' +-+ | .
. `-(::::::)-' | .
. | .
. .-(::::::::) | .
. .-(::::::::::::)-. +-+ | .
. (:::: Segment C :::)--|B|---+ .
. `-(::::::::::::)-' +-+ | .
. `-(::::::)-' | .
. | .
. ..(etc).. x .
. .
. .
. <- OMNI link Bridged by encapsulation -> .
. . . . . . . . . . . . . .. . . . . . . . .
Figure 2: Bridging OMNI Link Segments
Bridges, Servers, Relays and Proxys connect via secured INET tunnels
over their respective segments in a spanning tree topology rooted at
the Bridges. The secured spanning tree supports strong
authentication for IPv6 ND control messages and may also be used to
convey the initial data packets in a flow. Route optimization can
then be employed to cause data packets to take more direct paths
between OMNI link neighbors without having to strictly follow the
spanning tree.
3.2.2. Link-Local Addresses (LLAs) and Domain Local Addresses (DLAs)
AERO nodes on OMNI links use the Link-Local Address (LLA) prefix
fe80::/10 [RFC4291] to assign LLAs used for network-layer addresses
in IPv6 ND and data messages. They also use the Domain Local Address
(DLA) prefix [DLA]::/10 to form DLAs used for OAL header source and
destination addresses. See [I-D.templin-6man-omni-interface] for a
full specification of the LLAs and DLAs used by AERO nodes on OMNI
links.
Templin Expires May 6, 2021 [Page 14]
Internet-Draft AERO November 2020
IPv6 routers are not permitted to forward packets with LLA addresses.
Therefore, IPv6 packets with LLA addresses must be encapsulated in
IPv6 headers with DLA addresses in order to traverse the OMNI link.
DLA addresses therefore must be routable within the OMNI link domain-
local area.
For routing system organization (see Section 3.2.3), DLAs are
organized in partition prefixes, e.g., [DLA]::1000/116. For each
such partition prefix, the Bridge(s) that connect that segment assign
the all-zero's address of the prefix as a Subnet Router Anycast
address. For example, the Subnet Router Anycast address for
[DLA]::1000/116 is simply [DLA]::1000.
3.2.3. AERO Routing System
The AERO routing system comprises a private instance of the Border
Gateway Protocol (BGP) [RFC4271] that is coordinated between Bridges
and Servers and does not interact with either the public Internet BGP
routing system or any underlying INET routing systems.
In a reference deployment, each Server is configured as an Autonomous
System Border Router (ASBR) for a stub Autonomous System (AS) using
an AS Number (ASN) that is unique within the BGP instance, and each
Server further uses eBGP to peer with one or more Bridges but does
not peer with other Servers. Each INET of a multi-segment OMNI link
must include one or more Bridges, which peer with the Servers and
Proxys within that INET. All Bridges within the same INET are
members of the same hub AS using a common ASN, and use iBGP to
maintain a consistent view of all active MNPs currently in service.
The Bridges of different INETs peer with one another using eBGP.
Bridges advertise the OMNI link's MSPs and any non-MNP routes to each
of their Servers. This means that any aggregated non-MNPs (including
"default") are advertised to all Servers. Each Bridge configures a
black-hole route for each of its MSPs. By black-holing the MSPs, the
Bridge will maintain forwarding table entries only for the MNPs that
are currently active, and packets destined to all other MNPs will
correctly incur Destination Unreachable messages due to the black-
hole route. In this way, Servers have only partial topology
knowledge (i.e., they know only about the MNPs of their directly
associated Clients) and they forward all other packets to Bridges
which have full topology knowledge.
Each OMNI link segment assigns a unique sub-prefix of [DLA]::/96
known as the DLA partition prefix. For example, a first segment
could assign [DLA]::1000/116, a second could assign [DLA]::2000/116,
a third could assign [DLA]::3000/116, etc. The administrative
authorities for each segment must therefore coordinate to assure
Templin Expires May 6, 2021 [Page 15]
Internet-Draft AERO November 2020
mutually-exclusive partition prefix assignments, but internal
provisioning of each prefix is an independent local consideration for
each administrative authority.
DLA partition prefixes are statically represented in Bridge
forwarding tables. Bridges join multiple segments into a unified
OMNI link over multiple diverse administrative domains. They support
a bridging function by first establishing forwarding table entries
for their partition prefixes either via standard BGP routing or
static routes. For example, if three Bridges ('A', 'B' and 'C') from
different segments serviced [DLA]::1000/116, [DLA]::2000/116 and
[DLA]::3000/116 respectively, then the forwarding tables in each
Bridge are as follows:
A: [DLA]::1000/116->local, [DLA]::2000/116->B, [DLA]::3000/116->C
B: [DLA]::1000/116->A, [DLA]::2000/116->local, [DLA]::3000/116->C
C: [DLA]::1000/116->A, [DLA]::2000/116->B, [DLA]::3000/116->local
These forwarding table entries are permanent and never change, since
they correspond to fixed infrastructure elements in their respective
segments.
DLA Client prefixes are instead dynamically advertised in the AERO
routing system by Servers and Relays that provide service for their
corresponding MNPs. For example, if three Servers ('D', 'E' and 'F')
service the MNPs 2001:db8:1000:2000::/56, 2001:db8:3000:4000::/56 and
2001:db8:5000:6000::/56 then the routing system would include:
D: [DLA]:2001:db8:1000:2000::/72
E: [DLA]:2001:db8:3000:4000::/72
F: [DLA]:2001:db8:5000:6000::/72
A full discussion of the BGP-based routing system used by AERO is
found in [I-D.ietf-rtgwg-atn-bgp].
3.2.4. OMNI Link Encapsulation
With the Client and partition prefixes in place in Bridge forwarding
tables, control and data packets sent between AERO nodes in different
OMNI link segments can be carried over the spanning tree via mid-
layer encapsulations using the OMNI Adaptation Layer (OAL) header
and/or the OMNI Routing Header (ORH). OAL encapsulation is used for
all IPv4 packets, for IPv6 packets for which no route optimization
addressing information for the final segment is available, for
Templin Expires May 6, 2021 [Page 16]
Internet-Draft AERO November 2020
transportation of IPv6 ND messages with LLAs, and when fragmentation
is necessary (see Section 5 of [I-D.templin-6man-omni-interface]).
OAL encapsulation is based on Generic Packet Tunneling in IPv6
[RFC2473], with the inclusion of an ORH as an extension to the OAL
header if final segment link-layer address and/or final IPv6
destination address information is necessary.
The ORH is formatted as shown in Figure 3:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Next Header | Hdr Ext Len | Routing Type | Segments Left |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Reserved | SRT | FMT | LHS (bits 0 - 15) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| LHS (bits 16 - 31) | ~
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ~
~ Link Layer Address (L2ADDR) ~
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
~ IPv6 Destination Address ~
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
~ Null Padding ~
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 3: OMNI Routing Header (ORH) Format
In this format:
o Next Header identifies the type of header immediately following
the ORH.
o Hdr Ext Len is the length of the Routing header in 8-octet units,
not including the first 8 octets.
o Routing Type is set to TBD (see IANA Considerations).
o Segments Left is set to the value 1 if an IPv6 Destination Address
is included, or the value 0 if no IPv6 Destination Address is
included. If Segments Left encodes any other value, the entire
ORH is ignored and the next header is processed.
o Reserved is set to 0 on transmission and ignored on reception.
o SRT - a 5-bit Segment Routing Topology prefix length value that
(when added to 96) determines the prefix length to apply to the
DLA formed from concatenating [DLA*]::/96 with the 32 bit LHS MSID
Templin Expires May 6, 2021 [Page 17]
Internet-Draft AERO November 2020
value that follows. For example, the value 16 corresponds to the
prefix length 112.
o FMT - a 3-bit "Framework/Mode/Type" code corresponding to the
included Link Layer Address as follows:
* When the most significant bit (i.e., "Framework") is set to 0,
L2ADDR is the INET encapsulation address of a Server/Proxy;
otherwise, it is the address for the Source/Target itself.
* When the next most significant bit (i.e., "Mode") is set to 0,
the Source/Target L2ADDR is on the open INET; otherwise, it is
(likely) located behind a Network Address Translator (NAT).
* When the least significant bit (i.e., "Type") is set to 0,
L2ADDR includes a UDP Port Number followed by an IPv4 address;
else, a UDP Port Number followed by an IPv6 address.
o LHS - the 32 bit MSID of the Last Hop Server/Proxy on the path to
the target. When SRT and LHS are both set to 0, the LHS is
considered unspecified in this IPv6 ND message. When SRT is set
to 0 and LHS is non-zero, the prefix length is set to 128. SRT
and LHS provide guidance to the OMNI interface forwarding
algorithm. Specifically, if SRT/LHS is located in the local OMNI
link segment then the OMNI interface can encapsulate according to
FMT/L2ADDR; else, it must forward according to the OMNI link
spanning tree.
o Link Layer Address (L2ADDR) - Formatted according to FMT, and
identifies the link-layer address (i.e., the encapsulation
address) of the source/target. The UDP Port Number appears in the
first two octets and the IP address appears in the next 4 octets
for IPv4 or 16 octets for IPv6. The Port Number and IP address
are recorded in ones-compliment "obfuscated" form per [RFC4380].
The OMNI interface forwarding algorithm uses FMT/L2ADDR to
determine the encapsulation address for forwarding when SRT/LHS is
located in the local OMNI link segment.
o IPv6 Destination Address is a 16-octet IPv6 address included only
when Segments Left is 1, and omitted otherwise.
o Null Padding contains 0/4 zero-valued octets as necessary to pad
the ORH to an integral number of 8-octet units.
When an AERO node uses OAL encapsulation for a packet with IPv6
source address 2001:db8:1:2::1 and IPv6 destination address
2001:db8:1000:2000::1, and the LHS is not yet known, it sets the OAL
header source address to its own DLA address (e.g., [DLA]::1000:2000)
Templin Expires May 6, 2021 [Page 18]
Internet-Draft AERO November 2020
and sets the destination address to [DLA]:2001:db8:1000:2000::. The
node also includes an ORH header as an extension to the OAL header if
necessary, then fragments the mid-layer packet if necessary (note
that the ORH header need not include an IPv6 destination address,
since the address is already included in the encapsulated IPv6
packet). Next, the node encapsulates each resulting OAL packet in an
INET header with source address set to its own INET address (e.g.,
192.0.2.100) and destination set to the INET address of a Bridge
(e.g., 192.0.2.1). The encapsulation format in the above example is
shown in Figure 4:
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| INET Header |
| src = 192.0.2.100 |
| dst = 192.0.2.1 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| OAL Header |
| src = [DLA]::1000:2000 |
| dst=[DLA]:2001:db8:1000:2000::|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ORH Header |
| (no IPv6 destination included)|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Inner IP Header |
| src = 2001:db8:1:2::1 |
| dst = 2001:db8:1000:2000::1 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
~ ~
~ Inner Packet Body ~
~ ~
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 4: OAL/ORH Encapsulation
In this format, the inner IP header and packet body are the original
IP packet, the OAL header is an IPv6 header prepared according to
[RFC2473], the ORH is a Routing Header extension of the OAL header,
and the INET header is prepared as discussed in Section 3.6.
For IPv6 packets with non-link-local addresses for which the LHS is
known, the AERO node can avoid OAL encapsulation while inserting an
ORH as an extension to the existing IPv6 header of the packet. The
node further changes the IPv6 destination address to the Subnet
Router Anycast address corresponding to the LHS DLA address (for
example, for LHS [DLA]::1000:2000 and with SRT prefix length 16, the
Subnet Router Anycast address is [DLA]::1000:0000). When the node
Templin Expires May 6, 2021 [Page 19]
Internet-Draft AERO November 2020
changes the IPv6 destination address, it places the original
destination address in the ORH header so that it can be restored upon
OMNI link egress. The ORH-only encapsulation is shown in
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| INET Header |
| src = 192.0.2.100 |
| dst = 192.0.2.1 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Inner IP Header |
| src = 2001:db8:1:2::1 |
| dst = [DLA]::1000:0000 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ORH Header |
| dst = 2001:db8:1000:2000::1 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
~ ~
~ Inner Packet Body ~
~ ~
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 5: ORH-Only Encapsulation
This gives rise to a routing system that contains both Client prefix
routes that may change dynamically due to regional node mobility and
partition prefix routes that rarely if ever change. The Bridges can
therefore provide link-layer bridging by sending packets over the
spanning tree instead of network-layer routing according to MNP
routes. As a result, opportunities for packet loss due to node
mobility between different segments are mitigated.
In normal operations, IPv6 ND messages are conveyed over secured
paths between OMNI link neighbors so that specific Proxys, Servers or
Relays can be addressed without being subject to mobility events.
Conversely, only the first few packets destined to Clients need to
traverse secured paths until route optimization can determine a more
direct path.
3.2.5. Segment Routing Topologies (SRTs)
The 16-bit sub-prefixes of [DLA]::/10 identify up to 64 distinct
Segment Routing Topologies (SRTs). Each SRT is a mutually-exclusive
OMNI link overlay instance using a distinct set of DLAs, and emulates
a Virtual LAN (VLAN) service for the OMNI link. In some cases (e.g.,
when redundant topologies are needed for fault tolerance and
reliability) it may be beneficial to deploy multiple SRTs that act as
Templin Expires May 6, 2021 [Page 20]
Internet-Draft AERO November 2020
independent overlay instances. A communication failure in one
instance therefore will not affect communications in other instances.
Each SRT is identified by a distinct value in bits 10-15 of
[DLA]::10, i.e., as [DLA0]::/16, [DLA1]::/16, [DLA2]::/16, etc. This
document asserts that up to four SRTs provide a level of safety
sufficient for critical communications such as civil aviation. Each
SRT is designated with a color that identifies a different OMNI link
instance as follows:
o Red - corresponds to [DLA0]::/16
o Green - corresponds to [DLA1]::/16
o Blue-1 - corresponds to [DLA2]::/16
o Blue-2 - corresponds to [DLA3]::/16
o [DLA4]::/16 through [DLA63]::/16 are available for additional
SRTs.
Each OMNI interface is identified by a unqiue interface name (e.g.,
omni0, omni1, omni2, etc.) and assigns an anycast DLA corresponding
to its SRT prefix. For example, the anycast DLA for the Green SRT is
simply [DLA1]::. The anycast DLA is used for OMNI interface
determination in Safety-Based Multilink (SBM) as discussed in
[I-D.templin-6man-omni-interface]. Each OMNI interface further
applies Performance-Based Multilink (PBM) internally.
3.2.6. Segment Routing For OMNI Link Selection
An original IPv6 source can direct an IPv6 packet to an AERO node by
including a standard IPv6 Segment Routing Header (SRH) [RFC8754] with
the anycast DLA for the selected SRT as either the IPv6 destination
or as an intermediate hop within the SRH. This allows the original
source to determine the specific OMNI link topology a packet will
traverse when there may be multiple alternatives.
When the AERO node processes the SRH and forwards the packet to the
correct OMNI interface, the OMNI interface writes the next IPv6
address from the SRH into the IPv6 desination address and decrements
Segments Left. If decrementing would cause Segments Left to become
0, the OMNI interface deletes the SRH before forwarding. This form
of Segment Routing supports Safety-Based Multilink (SBM).
Templin Expires May 6, 2021 [Page 21]
Internet-Draft AERO November 2020
3.2.7. Segment Routing Within the OMNI Link
AERO node OMNI interfaces can insert an OAL and/or ORH header for
Segment Routing within the OMNI link to influence the paths of
packets destined to targets in remote segments without requiring all
packets to traverse strict spanning tree paths.
When an AERO node's OMNI interface has a packet to send to a target
discovered through route optimization located in the same OMNI link
segment, it encapsulates the packet in OAL/ORH headers if necessary
as discussed above. The node then uses the target's Link Layer
Address (L2ADDR) information for INET encapsulation.
When an AERO node's OMNI interface has a packet to send to a route
optimization target located in a remote OMNI link segment, it
encapsulates the packet in OAL/ORH headers as discussed above while
forwarding the packet to a Bridge with destination set to the Subnet
Router Anycast address for the final OMNI link segment.
When a Bridge receives a packet destined to its Subnet Router Anycast
address with an ORH with SRT/LHS values corresponding to the local
segment, it examines the L2ADDR according to FMT and removes the ORH
from the packet. If the packet did not include an OAL header, the
Bridge writes the saved IPv6 address from the ORH into the packet's
IPv6 destination address. If the packet includes an OAL header, the
Bridge instead writes the DLA of the LHS into the OAL header
destination if the LHS is available; else, it writes the DLA
corresponding to the inner packet IPv6 destination address. Next,
the Bridge encapsulates the packet in an INET header according to
L2ADDR and forwards the packet within the INET either to the LHS
Server/Proxy or directly to the destination itself. In this way, the
Bridge participates in route optimization to reduce traffic load and
suboptimal routing through strict spanning tree paths.
3.3. OMNI Interface Characteristics
OMNI interfaces are virtual interfaces configured over one or more
underlying interfaces classified as follows:
o INET interfaces connect to an INET either natively or through one
or several IPv4 Network Address Translators (NATs). Native INET
interfaces have global IP addresses that are reachable from any
INET correspondent. All Server, Relay and Bridge interfaces are
native interfaces, as are INET-facing interfaces of Proxys. NATed
INET interfaces connect to a private network behind one or more
NATs that provide INET access. Clients that are behind a NAT are
required to send periodic keepalive messages to keep NAT state
alive when there are no data packets flowing.
Templin Expires May 6, 2021 [Page 22]
Internet-Draft AERO November 2020
o ANET interfaces connect to an ANET that is separated from the open
INET by a Proxy. Proxys can actively issue control messages over
the INET on behalf of the Client to reduce ANET congestion.
o VPNed interfaces use security encapsulation over the INET to a
Virtual Private Network (VPN) server that also acts as a Server or
Proxy. Other than the link-layer encapsulation format, VPNed
interfaces behave the same as Direct interfaces.
o Direct interfaces connect a Client directly to a Server or Proxy
without crossing any ANET/INET paths. An example is a line-of-
sight link between a remote pilot and an unmanned aircraft. The
same Client considerations apply as for VPNed interfaces.
OMNI interfaces use OAL/ORH encapsulation as necessary as discussed
in Section 3.2.4. OMNI interfaces use link-layer encapsulation (see:
Section 3.6) to exchange packets with OMNI link neighbors over INET
or VPNed interfaces. OMNI interfaces do not use link-layer
encapsulation over ANET and Direct underlying interfaces.
OMNI interfaces maintain a neighbor cache for tracking per-neighbor
state the same as for any interface. OMNI interfaces use ND messages
including Router Solicitation (RS), Router Advertisement (RA),
Neighbor Solicitation (NS) and Neighbor Advertisement (NA) for
neighbor cache management.
OMNI interfaces send ND messages with an OMNI option formatted as
specified in [I-D.templin-6man-omni-interface]. The OMNI option
includes prefix registration information and Interface Attributes
containing link information parameters for the OMNI interface's
underlying interfaces. Each OMNI option may include multiple
Interface Attributes sub-options, each identified by an ifIndex
value.
A Client's OMNI interface may be configured over multiple underlying
interface connections. For example, common mobile handheld devices
have both wireless local area network ("WLAN") and cellular wireless
links. These links are often used "one at a time" with low-cost WLAN
preferred and highly-available cellular wireless as a standby, but a
simultaneous-use capability could provide benefits. In a more
complex example, aircraft frequently have many wireless data link
types (e.g. satellite-based, cellular, terrestrial, air-to-air
directional, etc.) with diverse performance and cost properties.
If a Client's multiple underlying interfaces are used "one at a time"
(i.e., all other interfaces are in standby mode while one interface
is active), then ND message OMNI options include only a single
Interface Attributes sub-option set to constant values. In that
Templin Expires May 6, 2021 [Page 23]
Internet-Draft AERO November 2020
case, the Client would appear to have a single interface but with a
dynamically changing link-layer address.
If the Client has multiple active underlying interfaces, then from
the perspective of ND it would appear to have multiple link-layer
addresses. In that case, ND message OMNI options MAY include
multiple Interface Attributes sub-options - each with values that
correspond to a specific interface. Every ND message need not
include Interface Attributes for all underlying interfaces; for any
attributes not included, the neighbor considers the status as
unchanged.
Bridge, Server and Proxy OMNI interfaces may be configured over one
or more secured tunnel interfaces. The OMNI interface configures
both an LLA and its corresponding DLA, while the underlying secured
tunnel interfaces are either unnumbered or configure the same DLA.
The OMNI interface encapsulates each IP packet in OAL/ORH headers and
presents the packet to the underlying secured tunnel interface.
Routing protocols such as BGP that run over the OMNI interface do not
employ OAL/ORH encapsulation, but rather present the routing protocol
messages directly to the underlying secured tunnels while using the
DLA as the source address. This distinction must be honored
consistently according to each node's configuration so that the IP
forwarding table will associate discovered IP routes with the correct
interface.
3.4. OMNI Interface Initialization
AERO Servers, Proxys and Clients configure OMNI interfaces as their
point of attachment to the OMNI link. AERO nodes assign the MSPs for
the link to their OMNI interfaces (i.e., as a "route-to-interface")
to ensure that packets with destination addresses covered by an MNP
not explicitly assigned to a non-OMNI interface are directed to the
OMNI interface.
OMNI interface initialization procedures for Servers, Proxys, Clients
and Bridges are discussed in the following sections.
3.4.1. AERO Server/Relay Behavior
When a Server enables an OMNI interface, it assigns an LLA/DLA
appropriate for the given OMNI link segment. The Server also
configures secured tunnels with one or more neighboring Bridges and
engages in a BGP routing protocol session with each Bridge.
The OMNI interface provides a single interface abstraction to the IP
layer, but internally comprises multiple secured tunnels as well as
an NBMA nexus for sending encapsulated data packets to OMNI interface
Templin Expires May 6, 2021 [Page 24]
Internet-Draft AERO November 2020
neighbors. The Server further configures a service to facilitate ND
exchanges with AERO Clients and manages per-Client neighbor cache
entries and IP forwarding table entries based on control message
exchanges.
Relays are simply Servers that run a dynamic routing protocol to
redistribute routes between the OMNI interface and INET/EUN
interfaces (see: Section 3.2.3). The Relay provisions MNPs to
networks on the INET/EUN interfaces (i.e., the same as a Client would
do) and advertises the MSP(s) for the OMNI link over the INET/EUN
interfaces. The Relay further provides an attachment point of the
OMNI link to a non-MNP-based global topology.
3.4.2. AERO Proxy Behavior
When a Proxy enables an OMNI interface, it assigns an LLA/DLA and
configures permanent neighbor cache entries the same as for Servers.
The Proxy also configures secured tunnels with one or more
neighboring Bridges and maintains per-Client neighbor cache entries
based on control message exchanges. Importantly Proxys are often
configured to act as Servers, and vice-versa.
3.4.3. AERO Client Behavior
When a Client enables an OMNI interface, it sends RS messages with ND
parameters over its underlying interfaces to a Server, which returns
an RA message with corresponding parameters. (The RS/RA messages may
pass through a Proxy in the case of a Client's ANET interface, or
through one or more NATs in the case of a Client's INET interface.)
3.4.4. AERO Bridge Behavior
AERO Bridges configure an OMNI interface and assign the DLA Subnet
Router Anycast address for each OMNI link segment they connect to.
Bridges configure secured tunnels with Servers, Proxys and other
Bridges; they also configure LLAs/DLAs and permanent neighbor cache
entries the same as Servers. Bridges engage in a BGP routing
protocol session with a subset of the Servers and other Bridges on
the spanning tree (see: Section 3.2.3).
3.5. OMNI Interface Neighbor Cache Maintenance
Each OMNI interface maintains a conceptual neighbor cache that
includes an entry for each neighbor it communicates with on the OMNI
link per [RFC4861]. OMNI interface neighbor cache entries are said
to be one of "permanent", "symmetric", "asymmetric" or "proxy".
Templin Expires May 6, 2021 [Page 25]
Internet-Draft AERO November 2020
Permanent neighbor cache entries are created through explicit
administrative action; they have no timeout values and remain in
place until explicitly deleted. AERO Bridges maintain permanent
neighbor cache entries for their associated Proxys/Servers (and vice-
versa). Each entry maintains the mapping between the neighbor's
network-layer LLA and corresponding INET address.
Symmetric neighbor cache entries are created and maintained through
RS/RA exchanges as specified in Section 3.12, and remain in place for
durations bounded by prefix lifetimes. AERO Servers maintain
symmetric neighbor cache entries for each of their associated
Clients, and AERO Clients maintain symmetric neighbor cache entries
for each of their associated Servers.
Asymmetric neighbor cache entries are created or updated based on
route optimization messaging as specified in Section 3.14, and are
garbage-collected when keepalive timers expire. AERO ROSs maintain
asymmetric neighbor cache entries for active targets with lifetimes
based on ND messaging constants. Asymmetric neighbor cache entries
are unidirectional since only the ROS (and not the ROR) creates an
entry.
Proxy neighbor cache entries are created and maintained by AERO
Proxys when they process Client/Server ND exchanges, and remain in
place for durations bounded by ND and prefix lifetimes. AERO Proxys
maintain proxy neighbor cache entries for each of their associated
Clients. Proxy neighbor cache entries track the Client state and the
address of the Client's associated Server(s).
To the list of neighbor cache entry states in Section 7.3.2 of
[RFC4861], Proxy and Server OMNI interfaces add an additional state
DEPARTED that applies to symmetric and proxy neighbor cache entries
for Clients that have recently departed. The interface sets a
"DepartTime" variable for the neighbor cache entry to "DEPART_TIME"
seconds. DepartTime is decremented unless a new ND message causes
the state to return to REACHABLE. While a neighbor cache entry is in
the DEPARTED state, packets destined to the target Client are
forwarded to the Client's new location instead of being dropped.
When DepartTime decrements to 0, the neighbor cache entry is deleted.
It is RECOMMENDED that DEPART_TIME be set to the default constant
value REACHABLE_TIME plus 10 seconds (40 seconds by default) to allow
a window for packets in flight to be delivered while stale route
optimization state may be present.
When an ROR receives an authentic NS message used for route
optimization, it searches for a symmetric neighbor cache entry for
the target Client. The ROR then returns a solicited NA message
without creating a neighbor cache entry for the ROS, but creates or
Templin Expires May 6, 2021 [Page 26]
Internet-Draft AERO November 2020
updates a target Client "Report List" entry for the ROS and sets a
"ReportTime" variable for the entry to REPORT_TIME seconds. The ROR
resets ReportTime when it receives a new authentic NS message, and
otherwise decrements ReportTime while no authentic NS messages have
been received. It is RECOMMENDED that REPORT_TIME be set to the
default constant value REACHABLE_TIME plus 10 seconds (40 seconds by
default) to allow a window for route optimization to converge before
ReportTime decrements below REACHABLE_TIME.
When the ROS receives a solicited NA message response to its NS
message used for route optimization, it creates or updates an
asymmetric neighbor cache entry for the target network-layer and
link-layer addresses. The ROS then (re)sets ReachableTime for the
neighbor cache entry to REACHABLE_TIME seconds and uses this value to
determine whether packets can be forwarded directly to the target,
i.e., instead of via a default route. The ROS otherwise decrements
ReachableTime while no further solicited NA messages arrive. It is
RECOMMENDED that REACHABLE_TIME be set to the default constant value
30 seconds as specified in [RFC4861].
AERO nodes also use the value MAX_UNICAST_SOLICIT to limit the number
of NS keepalives sent when a correspondent may have gone unreachable,
the value MAX_RTR_SOLICITATIONS to limit the number of RS messages
sent without receiving an RA and the value MAX_NEIGHBOR_ADVERTISEMENT
to limit the number of unsolicited NAs that can be sent based on a
single event. It is RECOMMENDED that MAX_UNICAST_SOLICIT,
MAX_RTR_SOLICITATIONS and MAX_NEIGHBOR_ADVERTISEMENT be set to 3 the
same as specified in [RFC4861].
Different values for DEPART_TIME, REPORT_TIME, REACHABLE_TIME,
MAX_UNICAST_SOLICIT, MAX_RTR_SOLCITATIONS and
MAX_NEIGHBOR_ADVERTISEMENT MAY be administratively set; however, if
different values are chosen, all nodes on the link MUST consistently
configure the same values. Most importantly, DEPART_TIME and
REPORT_TIME SHOULD be set to a value that is sufficiently longer than
REACHABLE_TIME to avoid packet loss due to stale route optimization
state.
3.5.1. OMNI Neighbor Interface Attributes
OMNI interface IPv6 ND messages include OMNI options
[I-D.templin-6man-omni-interface] with Interface Attributes that
provide Link-Layer Address and QoS Preference information for the
neighbor's underlying interfaces. This information is stored in the
neighbor cache and provides the basis for the forwarding algorithm
specified in Section 3.10. The information is cumulative and
reflects the union of the OMNI information from the most recent ND
Templin Expires May 6, 2021 [Page 27]
Internet-Draft AERO November 2020
messages received from the neighbor; it is therefore not required
that each ND message contain all neighbor information.
The OMNI option Interface Attributes for each underlying interface
includes a two-part "Link-Layer Address" consisting of a simple IP
encapsulation address determined by the FMT and L2ADDR fields and an
OAL DLA determined by the SRT and LHS fields. If the neighbor is
located in the local OMNI link segment (and, if any necessary NAT
state has been established) forwarding via simple IP encapsulation
can be used; otherwise, OAL encapsulation must be used. Underlying
interfaces are further selected based on their associated preference
values "high", "medium", "low" or "disabled".
Note: the OMNI option is distinct from any Source/Target Link-Layer
Address Options (S/TLLAOs) that may appear in an ND message according
to the appropriate IPv6 over specific link layer specification (e.g.,
[RFC2464]). If both an OMNI option and S/TLLAO appear, the former
pertains to encapsulation addresses while the latter pertains to the
native L2 address format of the underlying media.
3.5.2. OMNI Neighbor Advertisement Message Flags
As discussed in Section 4.4 of [RFC4861] NA messages include three
flag bits R, S and O. OMNI interface NA messages treat the flags as
follows:
o R: The R ("Router") flag is set to 1 in the NA messages sent by
all AERO/OMNI node types. Simple hosts that would set R to 0 do
not occur on the OMNI link itself, but may occur on the downstream
links of Clients and Relays.
o S: The S ("Solicited") flag is set exactly as specified in
Section 4.4. of [RFC4861], i.e., it is set to 1 for Solicited NAs
and set to 0 for Unsolicited NAs (both unicast and multicast).
o O: The O ("Override") flag is set to 0 for solicited proxy NAs and
set to 1 for all other solicited and unsolicited NAs. For further
study is whether solicited NAs for anycast targets apply for OMNI
links. Since OMNI LLAs must be uniquely assigned to Clients to
support correct ND protocol operation, however, no role is
currently seen for assigning the same OMNI LLA to multiple
Clients.
3.6. OMNI Interface Encapsulation and Re-encapsulation
The OMNI Adaptation Layer (OAL) inserts mid-layer IPv6 headers known
as the OAL/ORH headers when necessary as discussed in the following
sections. After either inserting or omitting the OAL/ORH headers,
Templin Expires May 6, 2021 [Page 28]
Internet-Draft AERO November 2020
the OMNI interface also inserts or omits an outer encapsulation
header as discussed below.
OMNI interfaces avoid outer encapsulation over Direct underlying
interfaces and ANET underlying interfaces for which the first-hop
access router is connected to the same underlying link. Otherwise,
OMNI interfaces encapsulate packets according to whether they are
entering the OMNI interface from the network layer or if they are
being re-admitted into the same OMNI link they arrived on. This
latter form of encapsulation is known as "re-encapsulation".
For packets entering the OMNI interface from the network layer, the
OMNI interface copies the "TTL/Hop Limit", "Type of Service/Traffic
Class" [RFC2983], "Flow Label"[RFC6438] (for IPv6) and "Congestion
Experienced" [RFC3168] values in the inner packet's IP header into
the corresponding fields in the OAL and outer encapsulation
header(s).
For packets undergoing re-encapsulation, the OMNI interface instead
copies these values from the original encapsulation header into the
new encapsulation header, i.e., the values are transferred between
encapsulation headers and *not* copied from the encapsulated packet's
network-layer header. (Note especially that by copying the TTL/Hop
Limit between encapsulation headers the value will eventually
decrement to 0 if there is a (temporary) routing loop.)
OMNI interfaces configured over ANET underlying interfaces which
employ a different IP protocol version and/or may be located multiple
IP hops from the nearest Proxy/Server use IP-in-IP encapsulation so
that the inner packet can traverse the ANET. IPv6 underlying ANET
interfaces use [RFC2473] encapsulation, while IPv4 interfaces use the
appropriate encapsulation per one of [RFC2529][RFC5214][RFC2003].
OMNI interfaces configured over INET underlying interfaces
encapsulate packets in INET headers according to the next hop
determined in the forwarding algorithm in Section 3.10. If the next
hop is reached via a secured tunnel, the OMNI interface uses an
encapsulation format specific to the secured tunnel type (see:
Section 6). If the next hop is reached via an unsecured INET
interface, the OMNI interface instead uses UDP/IP encapsulation per
[RFC4380] and as extended in [RFC6081].
When UDP/IP encapsulation is used, the OMNI interface next sets the
UDP source port to a constant value that it will use in each
successive packet it sends, and sets the UDP length field to the
length of the encapsulated packet plus 8 bytes for the UDP header
itself plus the length of any included extension headers or trailers.
Templin Expires May 6, 2021 [Page 29]
Internet-Draft AERO November 2020
The encapsulated packet may be either IPv6 or IPv4, as distinguished
by the version number found in the first four bits.
For UDP/IP-encapsulated packets sent to a Server, Relay or Bridge,
the OMNI interface sets the UDP destination port to 8060, i.e., the
IANA-registered port number for AERO. For packets sent to a Client,
the OMNI interface sets the UDP destination port to the port value
stored in the neighbor cache entry for this Client. The OMNI
interface finally includes/omits the UDP checksum according to
[RFC6935][RFC6936].
3.7. OMNI Interface Decapsulation
OMNI interfaces decapsulate packets destined either to the AERO node
itself or to a destination reached via an interface other than the
OMNI interface the packet was received on. When the encapsulated
packet arrives in multiple OAL fragments, the OMNI interface
reassembles as discussed in Section 3.9. Further decapsulation steps
are performed according to the appropriate encapsulation format
specification.
3.8. OMNI Interface Data Origin Authentication
AERO nodes employ simple data origin authentication procedures. In
particular:
o AERO Bridges, Servers and Proxys accept encapsulated data packets
and control messages received from the (secured) spanning tree.
o AERO Proxys and Clients accept packets that originate from within
the same secured ANET.
o AERO Clients and Relays accept packets from downstream network
correspondents based on ingress filtering.
o AERO Clients, Relays and Servers verify the outer UDP/IP
encapsulation addresses according to [RFC4380].
AERO nodes silently drop any packets that do not satisfy the above
data origin authentication procedures. Further security
considerations are discussed in Section 6.
3.9. OMNI Adaptation Layer and OMNI Interface MTU
The OMNI interface observes the link nature of tunnels, including the
Maximum Transmission Unit (MTU), Maximum Reassembly Unit (MRU) and
the role of fragmentation and reassembly [I-D.ietf-intarea-tunnels].
The OMNI interface employs an OMNI Adaptation Layer (OAL) for
Templin Expires May 6, 2021 [Page 30]
Internet-Draft AERO November 2020
accommodating multiple underlying links with diverse MTUs. The
functions of the OAL and the OMNI interface MTU/MRU are specified in
Section 5 of [I-D.templin-6man-omni-interface], with MTU/MRU both set
to the constant value 9180 bytes.
3.10. OMNI Interface Forwarding Algorithm
IP packets enter a node's OMNI interface either from the network
layer (i.e., from a local application or the IP forwarding system) or
from the link layer (i.e., from an OMNI interface neighbor). All
packets entering a node's OMNI interface first undergo data origin
authentication as discussed in Section 3.8. Packets that satisfy
data origin authentication are processed further, while all others
are dropped silently. The OMNI interface OAL wraps accepted packets
in OAL/ORH headers if necessary as discussed above.
Packets that enter the OMNI interface from the network layer are
forwarded to an OMNI interface neighbor. Packets that enter the OMNI
interface from the link layer are either re-admitted into the OMNI
link or forwarded to the network layer where they are subject to
either local delivery or IP forwarding. In all cases, the OMNI
interface itself MUST NOT decrement the network layer TTL/Hop-count
since its forwarding actions occur below the network layer.
OMNI interfaces may have multiple underlying interfaces and/or
neighbor cache entries for neighbors with multiple underlying
interfaces (see Section 3.3). The OMNI interface uses interface
attributes and/or traffic classifiers (e.g., DSCP value, port number,
etc.) to select an outgoing underlying interface for each packet
based on the node's own QoS preferences, and also to select a
destination link-layer address based on the neighbor's underlying
interface with the highest preference. AERO implementations SHOULD
allow for QoS preference values to be modified at runtime through
network management.
If multiple outgoing interfaces and/or neighbor interfaces have a
preference of "high", the AERO node replicates the packet and sends
one copy via each of the (outgoing / neighbor) interface pairs;
otherwise, the node sends a single copy of the packet via an
interface with the highest preference. AERO nodes keep track of
which underlying interfaces are currently "reachable" or
"unreachable", and only use "reachable" interfaces for forwarding
purposes.
The following sections discuss the OMNI interface forwarding
algorithms for Clients, Proxys, Servers and Bridges. In the
following discussion, a packet's destination address is said to
Templin Expires May 6, 2021 [Page 31]
Internet-Draft AERO November 2020
"match" if it is the same as a cached address, or if it is covered by
a cached prefix (which may be encoded in an LLA).
3.10.1. Client Forwarding Algorithm
When an IP packet enters a Client's OMNI interface from the network
layer the Client searches for an asymmetric neighbor cache entry that
matches the destination. If there is a match, the Client uses one or
more "reachable" neighbor interfaces in the entry for packet
forwarding. If there is no asymmetric neighbor cache entry, the
Client instead forwards the packet toward a Server (the packet is
intercepted by a Proxy if there is a Proxy on the path). The Client
encapsulates the packet in OAL/ORH headers if necessary and fragments
according to MTU requirements (see: Section 3.9).
When an IP packet enters a Client's OMNI interface from the link-
layer, if the destination matches one of the Client's MNPs or link-
local addresses the Client reassembles and decapsulates as necessary
and delivers the inner packet to the network layer. Otherwise, the
Client drops the packet and MAY return a network-layer ICMP
Destination Unreachable message subject to rate limiting (see:
Section 3.11).
3.10.2. Proxy Forwarding Algorithm
For control messages originating from or destined to a Client, the
Proxy intercepts the message and updates its proxy neighbor cache
entry for the Client. The Proxy then forwards a (proxyed) copy of
the control message. (For example, the Proxy forwards a proxied
version of a Client's NS/RS message to the target neighbor, and
forwards a proxied version of the NA/RA reply to the Client.)
When the Proxy receives a data packet from a Client within the ANET,
the Proxy reassembles and re-fragments if necessary then searches for
an asymmetric neighbor cache entry that matches the destination and
forwards as follows:
o if the destination matches an asymmetric neighbor cache entry, the
Proxy uses one or more "reachable" neighbor interfaces in the
entry for packet forwarding using OAL/ORH encapsulation if
necessary according to the cached link-layer address information.
If the neighbor interface is in the same OMNI link segment, the
Proxy forwards the packet directly to the neighbor; otherwise, it
forwards the packet to a Bridge.
o else, the Proxy uses OAL/ORH encapsulation and forwards the packet
to a Bridge while using the DLA corresponding to the packet's
destination as the destination address.
Templin Expires May 6, 2021 [Page 32]
Internet-Draft AERO November 2020
When the Proxy receives an encapsulated data packet from an INET
neighbor or from a secured tunnel from a Bridge, it accepts the
packet only if data origin authentication succeeds and if there is a
proxy neighbor cache entry that matches the inner destination. Next,
the Proxy reassembles the packet (if necessary) and continues
processing. If the reassembly is complete and the neighbor cache
state is REACHABLE, the Proxy then returns a PTB if necessary (see:
Section 3.9) then either drops or forwards the packet to the Client
while performing OAL/ORH encapsulation and re-fragmentation if
necessary. If the neighbor cache entry state is DEPARTED, the Proxy
instead changes the destination address to the address of the new
Server and forwards it to a Bridge while performing OAL/ORH re-
fragmentation if necessary.
3.10.3. Server/Relay Forwarding Algorithm
For control messages destined to a target Client's LLA that are
received from a secured tunnel, the Server intercepts the message and
sends a Proxyed response on behalf of the Client. (For example, the
Server sends a Proxyed NA message reply in response to an NS message
directed to one of its associated Clients.) If the Client's neighbor
cache entry is in the DEPARTED state, however, the Server instead
forwards the packet to the Client's new Server as discussed in
Section 3.16.
When the Server receives an encapsulated data packet from an INET
neighbor or from a secured tunnel, it accepts the packet only if data
origin authentication succeeds. The Server then continues processing
as follows:
o if the network layer destination matches a symmetric neighbor
cache entry in the REACHABLE state the Server prepares the packet
for forwarding to the destination Client. The Server first
reassembles (if necessary) and forwards the packet (while re-
fragmenting if necessary) as specified in Section 3.9.
o else, if the destination matches a symmetric neighbor cache entry
in the DEPARETED state the Server re-encapsulates the packet and
forwards it using the DLA of the Client's new Server as the
destination.
o else, if the destination matches an asymmetric neighbor cache
entry, the Server uses one or more "reachable" neighbor interfaces
in the entry for packet forwarding via the local INET if the
neighbor is in the same OMNI link segment or using OAL/ORH
encapsulation if necessary with the final destination set to the
neighbor's DLA otherwise.
Templin Expires May 6, 2021 [Page 33]
Internet-Draft AERO November 2020
o else, if the destination matches a non-MNP route in the IP
forwarding table or an LLA assigned to the Server's OMNI
interface, the Server reassembles if necessary, decapsulates the
packet and releases it to the network layer for local delivery or
IP forwarding.
o else, the Server drops the packet.
When the Server's OMNI interface receives a data packet from the
network layer or from a VPNed or Direct Client, it performs OAL/ORH
encapsulation and fragmentation if necessary, then processes the
packet according to the network-layer destination address as follows:
o if the destination matches a symmetric or asymmetric neighbor
cache entry the Server processes the packet as above.
o else, the Server encapsulates the packet in OLA/ORH headers and
forwards it to a Bridge using its own DLA as the source and the
DLA corresponding to the destination as the destination.
3.10.4. Bridge Forwarding Algorithm
Bridges forward OAL/ORH-encapsulated packets over secured tunnels the
same as any IP router. When the Bridge receives an OAL/ORH-
encapsulated packet via a secured tunnel, it removes the outer INET
header and searches for a forwarding table entry that matches the
destination address. The Bridge then processes the packet as
follows:
o if the destination matches its DLA Subnet Router Anycast address,
the Bridge checks for an ORH. If there is an ORH with SRT/LHS
located on the local segment, the Bridge removes the ORH from the
packet. (If there is no OAL header, the Bridge also writes the
saved IPv6 destination address from the ORH into the packet's IPv6
destination address. If there is an OAL header, the Bridge
instead writes the DLA of the LHS or the DLA corresponding to the
packet's IPv6 destination address into the OAL destination
address.) Next, the Bridge examines the FMT to determine if the
target is behind a NAT. If no NAT is indicated, the Bridge
forwards the packet directly to the L2ADDR using link-layer (UDP/
IP) encapsulation. If a NAT is indicated, the Bridge MAY perform
NAT traversal procedures by sending bubbles per [RFC4380]. The
Bridge then either applies AERO route optimization if NAT
traversal procedures have been successfully applied, or forwards
the packet directly to the Server indicated by the SR/LHS values.
o if the destination matches one of the Bridge's own addresses, the
Bridge submits the packet for local delivery.
Templin Expires May 6, 2021 [Page 34]
Internet-Draft AERO November 2020
o else, if the destination matches a forwarding table entry the
Bridge forwards the packet via a secured tunnel to the next hop.
If the destination matches an MSP without matching an MNP,
however, the Bridge instead drops the packet and returns an ICMP
Destination Unreachable message subject to rate limiting (see:
Section 3.11).
o else, the Bridge drops the packet and returns an ICMP Destination
Unreachable as above.
As for any IP router, the Bridge decrements the TTL/Hop Limit when it
forwards the packet. Therefore, when an OAL header is present only
the Hop Limit in the OAL header is decremented and not the TTL/Hop
Limit in the inner packet header. Bridges do not insert OAL/ORH
headers themselves; instead, they act as IPv6 routers and forward
packets based on the destination address found in the headers of
packets they receive.
3.11. OMNI Interface Error Handling
When an AERO node admits a packet into the OMNI interface, it may
receive link-layer or network-layer error indications.
A link-layer error indication is an ICMP error message generated by a
router in the INET on the path to the neighbor or by the neighbor
itself. The message includes an IP header with the address of the
node that generated the error as the source address and with the
link-layer address of the AERO node as the destination address.
The IP header is followed by an ICMP header that includes an error
Type, Code and Checksum. Valid type values include "Destination
Unreachable", "Time Exceeded" and "Parameter Problem"
[RFC0792][RFC4443]. (OMNI interfaces ignore all link-layer IPv4
"Fragmentation Needed" and IPv6 "Packet Too Big" messages since they
only emit packets that are guaranteed to be no larger than the IP
minimum link MTU as discussed in Section 3.9.)
The ICMP header is followed by the leading portion of the packet that
generated the error, also known as the "packet-in-error". For
ICMPv6, [RFC4443] specifies that the packet-in-error includes: "As
much of invoking packet as possible without the ICMPv6 packet
exceeding the minimum IPv6 MTU" (i.e., no more than 1280 bytes). For
ICMPv4, [RFC0792] specifies that the packet-in-error includes:
"Internet Header + 64 bits of Original Data Datagram", however
[RFC1812] Section 4.3.2.3 updates this specification by stating: "the
ICMP datagram SHOULD contain as much of the original datagram as
possible without the length of the ICMP datagram exceeding 576
bytes".
Templin Expires May 6, 2021 [Page 35]
Internet-Draft AERO November 2020
The link-layer error message format is shown in Figure 6 (where, "L2"
and "L3" refer to link-layer and network-layer, respectively):
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
~ ~
| L2 IP Header of |
| error message |
~ ~
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| L2 ICMP Header |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---
~ ~ P
| IP and other encapsulation | a
| headers of original L3 packet | c
~ ~ k
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ e
~ ~ t
| IP header of |
| original L3 packet | i
~ ~ n
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
~ ~ e
| Upper layer headers and | r
| leading portion of body | r
| of the original L3 packet | o
~ ~ r
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ---
Figure 6: OMNI Interface Link-Layer Error Message Format
The AERO node rules for processing these link-layer error messages
are as follows:
o When an AERO node receives a link-layer Parameter Problem message,
it processes the message the same as described as for ordinary
ICMP errors in the normative references [RFC0792][RFC4443].
o When an AERO node receives persistent link-layer Time Exceeded
messages, the IP ID field may be wrapping before earlier fragments
awaiting reassembly have been processed. In that case, the node
should begin including integrity checks and/or institute rate
limits for subsequent packets.
o When an AERO node receives persistent link-layer Destination
Unreachable messages in response to encapsulated packets that it
sends to one of its asymmetric neighbor correspondents, the node
should process the message as an indication that a path may be
failing, and optionally initiate NUD over that path. If it
Templin Expires May 6, 2021 [Page 36]
Internet-Draft AERO November 2020
receives Destination Unreachable messages over multiple paths, the
node should allow future packets destined to the correspondent to
flow through a default route and re-initiate route optimization.
o When an AERO Client receives persistent link-layer Destination
Unreachable messages in response to encapsulated packets that it
sends to one of its symmetric neighbor Servers, the Client should
mark the path as unusable and use another path. If it receives
Destination Unreachable messages on many or all paths, the Client
should associate with a new Server and release its association
with the old Server as specified in Section 3.16.5.
o When an AERO Server receives persistent link-layer Destination
Unreachable messages in response to encapsulated packets that it
sends to one of its symmetric neighbor Clients, the Server should
mark the underlying path as unusable and use another underlying
path.
o When an AERO Server or Proxy receives link-layer Destination
Unreachable messages in response to an encapsulated packet that it
sends to one of its permanent neighbors, it treats the messages as
an indication that the path to the neighbor may be failing.
However, the dynamic routing protocol should soon reconverge and
correct the temporary outage.
When an AERO Bridge receives a packet for which the network-layer
destination address is covered by an MSP, if there is no more-
specific routing information for the destination the Bridge drops the
packet and returns a network-layer Destination Unreachable message
subject to rate limiting. The Bridge writes the network-layer source
address of the original packet as the destination address and uses
one of its non link-local addresses as the source address of the
message.
When an AERO node receives an encapsulated packet for which the
reassembly buffer it too small, it drops the packet and returns a
network-layer Packet Too Big (PTB) message. The node first writes
the MRU value into the PTB message MTU field, writes the network-
layer source address of the original packet as the destination
address and writes one of its non link-local addresses as the source
address.
3.12. AERO Router Discovery, Prefix Delegation and Autoconfiguration
AERO Router Discovery, Prefix Delegation and Autoconfiguration are
coordinated as discussed in the following Sections.
Templin Expires May 6, 2021 [Page 37]
Internet-Draft AERO November 2020
3.12.1. AERO Service Model
Each AERO Server on the OMNI link is configured to facilitate Client
prefix delegation/registration requests. Each Server is provisioned
with a database of MNP-to-Client ID mappings for all Clients enrolled
in the AERO service, as well as any information necessary to
authenticate each Client. The Client database is maintained by a
central administrative authority for the OMNI link and securely
distributed to all Servers, e.g., via the Lightweight Directory
Access Protocol (LDAP) [RFC4511], via static configuration, etc.
Clients receive the same service regardless of the Servers they
select.
AERO Clients and Servers use ND messages to maintain neighbor cache
entries. AERO Servers configure their OMNI interfaces as advertising
NBMA interfaces, and therefore send unicast RA messages with a short
Router Lifetime value (e.g., ReachableTime seconds) in response to a
Client's RS message. Thereafter, Clients send additional RS messages
to keep Server state alive.
AERO Clients and Servers include prefix delegation and/or
registration parameters in RS/RA messages (see
[I-D.templin-6man-omni-interface]). The ND messages are exchanged
between Client and Server according to the prefix management schedule
required by the service. If the Client knows its MNP in advance, it
can employ prefix registration by including its LLA as the source
address of an RS message and with an OMNI option with valid prefix
registration information for the MNP. If the Server (and Proxy)
accept the Client's MNP assertion, they inject the prefix into the
routing system and establish the necessary neighbor cache state.
The following sections specify the Client and Server behavior.
3.12.2. AERO Client Behavior
AERO Clients discover the addresses of Servers in a similar manner as
described in [RFC5214]. Discovery methods include static
configuration (e.g., from a flat-file map of Server addresses and
locations), or through an automated means such as Domain Name System
(DNS) name resolution [RFC1035]. Alternatively, the Client can
discover Server addresses through a layer 2 data link login exchange,
or through a unicast RA response to a multicast/anycast RS as
described below. In the absence of other information, the Client can
resolve the DNS Fully-Qualified Domain Name (FQDN)
"linkupnetworks.[domainname]" where "linkupnetworks" is a constant
text string and "[domainname]" is a DNS suffix for the OMNI link
(e.g., "example.com").
Templin Expires May 6, 2021 [Page 38]
Internet-Draft AERO November 2020
To associate with a Server, the Client acts as a requesting router to
request MNPs. The Client prepares an RS message with prefix
management parameters and includes a Nonce and Timestamp option if
the Client needs to correlate RA replies. If the Client already
knows the Server's LLA, it includes the LLA as the network-layer
destination address; otherwise, it includes (link-local) All-Routers
multicast as the network-layer destination. If the Client already
knows its own LLA, it uses the LLA as the network-layer source
address; otherwise, it uses an OMNI Temporary LLA as the network-
layer source address and includes a DHCP Unique Identifier (DUID)
sub-option in the OMNI option (see:
[I-D.templin-6man-omni-interface]).
The Client next includes an OMNI option in the RS message to register
its link-layer information with the Server. The Client sets the OMNI
option prefix registration information according to the MNP, and
includes Interface Attributes corresponding to the underlying
interface over which the Client will send the RS message. The Client
MAY include additional Interface Attributes specific to other
underlying interfaces.
The Client then sends the RS message (either directly via Direct
interfaces, via a VPN for VPNed interfaces, via a Proxy for ANET
interfaces or via INET encapsulation for INET interfaces) and waits
for an RA message reply (see Section 3.12.3). The Client retries up
to MAX_RTR_SOLICITATIONS times until an RA is received. If the
Client receives no RAs, or if it receives an RA with Router Lifetime
set to 0, the Client SHOULD abandon this Server and try another
Server. Otherwise, the Client processes the prefix information found
in the RA message.
Next, the Client creates a symmetric neighbor cache entry with the
Server's LLA as the network-layer address and the Server's
encapsulation and/or link-layer addresses as the link-layer address.
The Client records the RA Router Lifetime field value in the neighbor
cache entry as the time for which the Server has committed to
maintaining the MNP in the routing system via this underlying
interface, and caches the other RA configuration information
including Cur Hop Limit, M and O flags, Reachable Time and Retrans
Timer. The Client then autoconfigures LLAs for each of the delegated
MNPs and assigns them to the OMNI interface. The Client also caches
any MSPs included in Route Information Options (RIOs) [RFC4191] as
MSPs to associate with the OMNI link, and assigns the MTU value in
the MTU option to the underlying interface.
The Client then registers additional underlying interfaces with the
Server by sending RS messages via each additional interface. The RS
Templin Expires May 6, 2021 [Page 39]
Internet-Draft AERO November 2020
messages include the same parameters as for the initial RS/RA
exchange, but with destination address set to the Server's LLA.
Following autoconfiguration, the Client sub-delegates the MNPs to its
attached EUNs and/or the Client's own internal virtual interfaces as
described in [I-D.templin-v6ops-pdhost] to support the Client's
downstream attached "Internet of Things (IoT)". The Client
subsequently sends additional RS messages over each underlying
interface before the Router Lifetime received for that interface
expires.
After the Client registers its underlying interfaces, it may wish to
change one or more registrations, e.g., if an interface changes
address or becomes unavailable, if QoS preferences change, etc. To
do so, the Client prepares an RS message to send over any available
underlying interface. The RS includes an OMNI option with prefix
registration information specific to its MNP, with Interface
Attributes specific to the selected underlying interface, and with
any additional Interface Attributes specific to other underlying
interfaces. When the Client receives the Server's RA response, it
has assurance that the Server has been updated with the new
information.
If the Client wishes to discontinue use of a Server it issues an RS
message over any underlying interface with an OMNI option with a
prefix release indication. When the Server processes the message, it
releases the MNP, sets the symmetric neighbor cache entry state for
the Client to DEPARTED and returns an RA reply with Router Lifetime
set to 0. After a short delay (e.g., 2 seconds), the Server
withdraws the MNP from the routing system.
3.12.3. AERO Server Behavior
AERO Servers act as IP routers and support a prefix delegation/
registration service for Clients. Servers arrange to add their LLAs
to a static map of Server addresses for the link and/or the DNS
resource records for the FQDN "linkupnetworks.[domainname]" before
entering service. Server addresses should be geographically and/or
topologically referenced, and made available for discovery by Clients
on the OMNI link.
When a Server receives a prospective Client's RS message on its OMNI
interface, it SHOULD return an immediate RA reply with Router
Lifetime set to 0 if it is currently too busy or otherwise unable to
service the Client. Otherwise, the Server authenticates the RS
message and processes the prefix delegation/registration parameters.
The Server first determines the correct MNPs to provide to the Client
by searching the Client database. When the Server returns the MNPs,
Templin Expires May 6, 2021 [Page 40]
Internet-Draft AERO November 2020
it also creates a forwarding table entry for the DLA corresponding to
each MNP so that the MNPs are propagated into the routing system
(see: Section 3.2.3). For IPv6, the Server creates an IPv6
forwarding table entry for each MNP. For IPv4, the Server creates an
IPv6 forwarding table entry with the IPv4-compatibility DLA prefix
corresponding to the IPv4 address.
The Server next creates a symmetric neighbor cache entry for the
Client using the base LLA as the network-layer address and with
lifetime set to no more than the smallest prefix lifetime. Next, the
Server updates the neighbor cache entry by recording the information
in each Interface Attributes sub-option in the RS OMNI option. The
Server also records the actual OAL/INET addresses in the neighbor
cache entry.
Next, the Server prepares an RA message using its LLA as the network-
layer source address and the network-layer source address of the RS
message as the network-layer destination address. The Server sets
the Router Lifetime to the time for which it will maintain both this
underlying interface individually and the symmetric neighbor cache
entry as a whole. The Server also sets Cur Hop Limit, M and O flags,
Reachable Time and Retrans Timer to values appropriate for the OMNI
link. The Server includes the MNPs, any other prefix management
parameters and an OMNI option with no Interface Attributes. The
Server then includes one or more RIOs that encode the MSPs for the
OMNI link, plus an MTU option (see Section 3.9). The Server finally
forwards the message to the Client using OAL/INET, INET, or NULL
encapsulation as necessary.
After the initial RS/RA exchange, the Server maintains a
ReachableTime timer for each of the Client's underlying interfaces
individually (and for the Client's symmetric neighbor cache entry
collectively) set to expire after ReachableTime seconds. If the
Client (or Proxy) issues additional RS messages, the Server sends an
RA response and resets ReachableTime. If the Server receives an ND
message with a prefix release indication it sets the Client's
symmetric neighbor cache entry to the DEPARTED state and withdraws
the MNP from the routing system after a short delay (e.g., 2
seconds). If ReachableTime expires before a new RS is received on an
individual underlying interface, the Server marks the interface as
DOWN. If ReachableTime expires before any new RS is received on any
individual underlying interface, the Server sets the symmetric
neighbor cache entry state to STALE and sets a 10 second timer. If
the Server has not received a new RS or ND message with a prefix
release indication before the 10 second timer expires, it deletes the
neighbor cache entry and withdraws the MNP from the routing system.
Templin Expires May 6, 2021 [Page 41]
Internet-Draft AERO November 2020
The Server processes any ND messages pertaining to the Client and
returns an NA/RA reply in response to solicitations. The Server may
also issue unsolicited RA messages, e.g., with reconfigure parameters
to cause the Client to renegotiate its prefix delegation/
registrations, with Router Lifetime set to 0 if it can no longer
service this Client, etc. Finally, If the symmetric neighbor cache
entry is in the DEPARTED state, the Server deletes the entry after
DepartTime expires.
Note: Clients SHOULD notify former Servers of their departures, but
Servers are responsible for expiring neighbor cache entries and
withdrawing routes even if no departure notification is received
(e.g., if the Client leaves the network unexpectedly). Servers
SHOULD therefore set Router Lifetime to ReachableTime seconds in
solicited RA messages to minimize persistent stale cache information
in the absence of Client departure notifications. A short Router
Lifetime also ensures that proactive Client/Server RS/RA messaging
will keep any NAT state alive (see above).
Note: All Servers on an OMNI link MUST advertise consistent values in
the RA Cur Hop Limit, M and O flags, Reachable Time and Retrans Timer
fields the same as for any link, since unpredictable behavior could
result if different Servers on the same link advertised different
values.
3.12.3.1. DHCPv6-Based Prefix Registration
When a Client is not pre-provisioned with an OMNI LLA containing a
MNP, it will need for the Server to select an MNP on its behalf and
set up the correct state in the AERO routing service. The DHCPv6
service [RFC8415] is used to support this requirement.
When a Client without a pre-provisioned OMNI MN LLA needs to have the
Server select an MNP, it sends an RS message with source address set
to an OMNI Temporary LLA. The RS message includes an OMNI option
with a DUID sub-option that contains a unique DUID value for the MN,
and a Prefix Length value corresponding to the length of MNP it
wishes to receive.
When the Server receives the RS message, it notes that the source was
a Temporary LLA and derives the DUID and Prefix Length from the OMNI
option. The Server (acting as a "proxy DHCP client") then crafts a
well-formed DHCPv6 Solicit message with Prefix Delegation (PD)
parameters, and forwards the message to a locally-resident DHCPv6
server. The DHCPv6 server then delegates an MNP and returns a DHCPv6
Reply message with PD parameters. (Note: the Solicit message should
contain enough identifying information for the MN so that the AR can
correlate the server's Reply with the original Solicit.)
Templin Expires May 6, 2021 [Page 42]
Internet-Draft AERO November 2020
When the Server receives the DHCPv6 Reply, it adds a route to the
routing system and creates an OMNI MN LLA based on the delegated MNP.
The Server then sends an RA back to the Client with the (newly-
created) OMNI MN LLA as the destination address and with Prefix
Length set in the OMNI option. When the Client receives the RA, it
creates a default route, assigns the Subnet Router Anycast address
and sets its OMNI LLA based on the delegated MNP.
3.13. The AERO Proxy
Clients may connect to protected-spectrum ANETs that employ physical
and/or link-layer security services to facilitate communications to
Servers in outside INETs. In that case, the ANET can employ an AERO
Proxy. The Proxy is located at the ANET/INET border and listens for
RS messages originating from or RA messages destined to ANET Clients.
The Proxy acts on these control messages as follows:
o when the Proxy receives an RS message from a new ANET Client, it
first authenticates the message then examines the network-layer
destination address. If the destination address is a Server's
LLA, the Proxy proceeds to the next step. Otherwise, if the
destination is (link-local) All-Routers multicast, the Proxy
selects a "nearby" Server that is likely to be a good candidate to
serve the Client and replaces the destination address with the
Server's LLA. Next, the Proxy creates a proxy neighbor cache
entry and caches the Client and Server link-layer addresses along
with the OMNI option information and any other identifying
information including Transaction IDs, Client Identifiers, Nonce
values, etc. The Proxy finally encapsulates the (proxyed) RS
message in an OAL header with source set to the Proxy's DLA and
destination set to the Server's DLA. The Proxy also includes an
OMNI header with an Interface Attributes option that includes its
own INET address plus a unique Port Number for this Client, then
forwards the message into the OMNI link spanning tree.
o when the Server receives the RS, it authenticates the message then
creates or updates a symmetric neighbor cache entry for the Client
with the Proxy's DLA, INET address and Port Number as the link-
layer address information. The Server then sends an RA message
back to the Proxy via the spanning tree.
o when the Proxy receives the RA, it authenticates the message and
matches it with the proxy neighbor cache entry created by the RS.
The Proxy then caches the prefix information as a mapping from the
Client's MNPs to the Client's link-layer address, caches the
Server's advertised Router Lifetime and sets the neighbor cache
entry state to REACHABLE. The Proxy then optionally rewrites the
Router Lifetime and forwards the (proxyed) message to the Client.
Templin Expires May 6, 2021 [Page 43]
Internet-Draft AERO November 2020
The Proxy finally includes an MTU option (if necessary) with an
MTU to use for the underlying ANET interface.
After the initial RS/RA exchange, the Proxy forwards any Client data
packets for which there is no matching asymmetric neighbor cache
entry to a Bridge using OAL encapsulation with its own DLA as the
source and the DLA corresponding to the Client as the destination.
The Proxy instead forwards any Client data destined to an asymmetric
neighbor cache target directly to the target according to the OAL/
link-layer information - the process of establishing asymmetric
neighbor cache entries is specified in Section 3.14.
While the Client is still attached to the ANET, the Proxy sends NS,
RS and/or unsolicited NA messages to update the Server's symmetric
neighbor cache entries on behalf of the Client and/or to convey QoS
updates. This allows for higher-frequency Proxy-initiated RS/RA
messaging over well-connected INET infrastructure supplemented by
lower-frequency Client-initiated RS/RA messaging over constrained
ANET data links.
If the Server ceases to send solicited advertisements, the Proxy
sends unsolicited RAs on the ANET interface with destination set to
(link-local) All-Nodes multicast and with Router Lifetime set to zero
to inform Clients that the Server has failed. Although the Proxy
engages in ND exchanges on behalf of the Client, the Client can also
send ND messages on its own behalf, e.g., if it is in a better
position than the Proxy to convey QoS changes, etc. For this reason,
the Proxy marks any Client-originated solicitation messages (e.g. by
inserting a Nonce option) so that it can return the solicited
advertisement to the Client instead of processing it locally.
If the Client becomes unreachable, the Proxy sets the neighbor cache
entry state to DEPARTED and retains the entry for DepartTime seconds.
While the state is DEPARTED, the Proxy forwards any packets destined
to the Client to a Bridge via OAL encapsulation with the Client's
current Server as the destination. The Bridge in turn forwards the
packets to the Client's current Server. When DepartTime expires, the
Proxy deletes the neighbor cache entry and discards any further
packets destined to this (now forgotten) Client.
In some ANETs that employ a Proxy, the Client's MNP can be injected
into the ANET routing system. In that case, the Client can send data
messages without encapsulation so that the ANET routing system
transports the unencapsulated packets to the Proxy. This can be very
beneficial, e.g., if the Client connects to the ANET via low-end data
links such as some aviation wireless links.
Templin Expires May 6, 2021 [Page 44]
Internet-Draft AERO November 2020
If the first-hop ANET access router is on the same underlying link
and recognizes the AERO/OMNI protocol, the Client can avoid
encapsulation for both its control and data messages. When the
Client connects to the link, it can send an unencapsulated RS message
with source address set to its LLA and with destination address set
to the LLA of the Client's selected Server or to (link-local) All-
Routers multicast. The Client includes an OMNI option formatted as
specified in [I-D.templin-6man-omni-interface].
The Client then sends the unencapsulated RS message, which will be
intercepted by the AERO-Aware access router. The access router then
encapsulates the RS message in an ANET header with its own address as
the source address and the address of a Proxy as the destination
address. The access router further remembers the address of the
Proxy so that it can encapsulate future data packets from the Client
via the same Proxy. If the access router needs to change to a new
Proxy, it simply sends another RS message toward the Server via the
new Proxy on behalf of the Client.
In some cases, the access router and Proxy may be one and the same
node. In that case, the node would be located on the same physical
link as the Client, but its message exchanges with the Server would
need to pass through a security gateway at the ANET/INET border. The
method for deploying access routers and Proxys (i.e. as a single node
or multiple nodes) is an ANET-local administrative consideration.
3.13.1. Combined Proxy/Servers
Clients may need to connect directly to Servers via INET, Direct and
VPNed interfaces (i.e., non-ANET interfaces). If the Client's
underlying interfaces all connect via the same INET partition, then
it can connect to a single controlling Server via all interfaces.
If some Client interfaces connect via different INET partitions,
however, the Client still selects a set of controlling Servers and
sends RS messages via their directly-connected Servers while using
the LLA of the controlling Server as the destination.
When a Server receives an RS with destination set to the LLA of a
controlling Server, it acts as a Proxy to forward the message to the
controlling Server while forwarding the corresponding RA reply to the
Client.
3.13.2. Detecting and Responding to Server Failures
In environments where fast recovery from Server failure is required,
Proxys SHOULD use proactive Neighbor Unreachability Detection (NUD)
to track Server reachability in a similar fashion as for
Templin Expires May 6, 2021 [Page 45]
Internet-Draft AERO November 2020
Bidirectional Forwarding Detection (BFD) [RFC5880]. Proxys can then
quickly detect and react to failures so that cached information is
re-established through alternate paths. The NUD control messaging is
carried only over well-connected ground domain networks (i.e., and
not low-end aeronautical radio links) and can therefore be tuned for
rapid response.
Proxys perform proactive NUD with Servers for which there are
currently active ANET Clients by sending continuous NS messages in
rapid succession, e.g., one message per second. The Proxy sends the
NS message via the spanning tree with the Proxy's LLA as the source
and the LLA of the Server as the destination. When the Proxy is also
sending RS messages to the Server on behalf of ANET Clients, the
resulting RA responses can be considered as equivalent hints of
forward progress. This means that the Proxy need not also send a
periodic NS if it has already sent an RS within the same period. If
the Server fails (i.e., if the Proxy ceases to receive
advertisements), the Proxy can quickly inform Clients by sending
multicast RA messages on the ANET interface.
The Proxy sends RA messages on the ANET interface with source address
set to the Server's address, destination address set to (link-local)
All-Nodes multicast, and Router Lifetime set to 0. The Proxy SHOULD
send MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small
delays [RFC4861]. Any Clients on the ANET that had been using the
failed Server will receive the RA messages and associate with a new
Server.
3.13.3. Point-to-Multipoint Server Coordination
In environments where Client messaging over ANETs is bandwidth-
limited and/or expensive, Clients can enlist the services of the
Proxy to coordinate with multiple Servers in a single RS/RA message
exchange. The Client can send a single RS message to (link-local)
All-Routers multicast that includes the ID's of multiple Servers in
MS-Register sub-options of the OMNI option.
When the Proxy receives the RS and processes the OMNI option, it
sends a separate RS to each MS-Register Server ID. When the Proxy
receives an RA, it can optionally return an immediate "singleton" RA
to the Client or record the Server's ID for inclusion in a pending
"aggregate" RA message. The Proxy can then return aggregate RA
messages to the Client including multiple Server IDs in order to
conserve bandwidth. Each RA includes a proper subset of the Server
IDs from the original RS message, and the Proxy must ensure that the
message contents of each RA are consistent with the information
received from the (aggregated) Servers.
Templin Expires May 6, 2021 [Page 46]
Internet-Draft AERO November 2020
Clients can thereafter employ efficient point-to-multipoint Server
coordination under the assistance of the Proxy to reduce the number
of messages sent over the ANET while enlisting the support of
multiple Servers for fault tolerance. Clients can further include
MS-Release sub-options in IPv6 ND messages to request the Proxy to
release from former Servers via the procedures discussed in
Section 3.16.5.
The OMNI interface specification [I-D.templin-6man-omni-interface]
provides further discussion of the Client/Proxy RS/RA messaging
involved in point-to-multipoint coordination.
3.14. AERO Route Optimization / Address Resolution
While data packets are flowing between a source and target node,
route optimization SHOULD be used. Route optimization is initiated
by the first eligible Route Optimization Source (ROS) closest to the
source as follows:
o For Clients on VPNed and Direct interfaces, the Server is the ROS.
o For Clients on ANET interfaces, the Proxy is the ROS.
o For Clients on INET interfaces, the Client itself is the ROS.
o For correspondent nodes on INET/EUN interfaces serviced by a
Relay, the Relay is the ROS.
The route optimization procedure is conducted between the ROS and the
target Server/Relay acting as a Route Optimization Responder (ROR) in
the same manner as for IPv6 ND Address Resolution and using the same
NS/NA messaging. The target may either be a MNP Client serviced by a
Server, or a non-MNP correspondent reachable via a Relay.
The procedures are specified in the following sections.
3.14.1. Route Optimization Initiation
While data packets are flowing from the source node toward a target
node, the ROS performs address resolution by sending an NS message
for Address Resolution (NS(AR)) to receive a solicited NA message
from the ROR. When the ROS sends an NS(AR), it includes:
o the LLA of the ROS as the source address.
o the data packet's destination as the Target Address.
Templin Expires May 6, 2021 [Page 47]
Internet-Draft AERO November 2020
o the Solicited-Node multicast address [RFC4291] formed from the
lower 24 bits of the data packet's destination as the destination
address, e.g., for 2001:db8:1:2::10:2000 the NS destination
address is ff02:0:0:0:0:1:ff10:2000.
The NS(AR) message includes an OMNI option with no Interface
Attributes, such that the target will not create a neighbor cache
entry. The Prefix Length in the OMNI option is set to the Prefix
Length associated with the ROS's LLA.
The ROS then encapsulates the NS(AR) message in an OAL header with
source set to its own DLA and destination set to the DLA
corresponding to the target, then sends the message into the spanning
tree without decrementing the network-layer TTL/Hop Limit field.
(When the ROS is a Client, it instead securely sends the NS(AR) per
Section 3.22.1 to one of its current Servers, which forwards the
message into the spanning tree on behalf of the Client. When the
Server forwards the NS(AR), it sets the IPv6 source address and the
OAL source address to the LLA and DLA of the Client, respectively.)
3.14.2. Relaying the NS
When the Bridge receives the NS(AR) message from the ROS, it discards
the INET header and determines that the ROR is the next hop by
consulting its standard IPv6 forwarding table for the OAL header
destination address. The Bridge then forwards the message toward the
ROR via the spanning tree the same as for any IPv6 router. The
final-hop Bridge in the spanning tree will deliver the message via a
secured tunnel to the ROR.
3.14.3. Processing the NS and Sending the NA
When the ROR receives the NS(AR) message, it examines the Target
Address to determine whether it has a neighbor cache entry and/or
route that matches the target. If there is no match, the ROR drops
the message. Otherwise, the ROR continues processing as follows:
o if the target belongs to an MNP Client neighbor in the DEPARTED
state the ROR changes the NS(AR) message OAL destination address
to the DLA of the Client's new Server, forwards the message into
the spanning tree and returns from processing.
o If the target belongs to an MNP Client neighbor in the REACHABLE
state, the ROR instead adds the AERO source address to the target
Client's Report List with time set to ReportTime.
o If the target belongs to a non-MNP route, the ROR continues
processing without adding an entry to the Report List.
Templin Expires May 6, 2021 [Page 48]
Internet-Draft AERO November 2020
The ROR then prepares a solicited NA message to send back to the ROS
but does not create a neighbor cache entry. The ROR sets the NA
source address to the LLA corresponding to the target, sets the
Target Address to the target of the solicitation, and sets the
destination address to the source of the solicitation. The ROR then
includes an OMNI option with Prefix Length set to the length
associated with the LLA.
If the target is an MNP Client, the ROR next includes Interface
Attributes in the OMNI option for each of the target Client's
underlying interfaces with current information for each interface and
with the S/T-ifIndex field in the OMNI header set to 0 to indicate
that the message originated from the ROR and not the Client.
For each Interface Attributes sub-option, the ROR sets the L2ADDR
according to its own INET address for VPNed or Direct interfaces, to
the INET address of the Proxy or to the Client's INET address for
INET interfaces. The ROR then includes the lower 32 bits of its own
DLA (or the DLA of the Proxy) as the LHS, encodes the DLA prefix
length code in the SRT field and sets the FMT code accordingly as
specified in Section 3.3.
The ROR then sets the NA message R flag to 1 (as a router), S flag to
1 (as a response to a solicitation), and O flag to 0 (as a proxy).
The ROR finally encapsulates the NA message in an OAL header with
source set to its own DLA and destination set to the source DLA of
the NS(AR) message, then forwards the message into the spanning tree
without decrementing the network-layer TTL/Hop Limit field.
3.14.4. Relaying the NA
When the Bridge receives the NA message from the ROR, it discards the
INET header and determines that the ROS is the next hop by consulting
its standard IPv6 forwarding table for the OAL header destination
address. The Bridge then forwards the OAL-encapsulated NA message
toward the ROS the same as for any IPv6 router. The final-hop Bridge
in the spanning tree will deliver the message via a secured tunnel to
the ROS.
3.14.5. Processing the NA
When the ROS receives the solicited NA message, it processes the
message the same as for standard IPv6 Address Resolution [RFC4861].
In the process, it caches the source DLA then creates an asymmetric
neighbor cache entry for the target and caches all information found
in the OMNI option. The ROS finally sets the asymmetric neighbor
cache entry lifetime to ReachableTime seconds. (When the ROS is a
Client, the solicited NA message will first be delivered via the
Templin Expires May 6, 2021 [Page 49]
Internet-Draft AERO November 2020
spanning tree to one of its current Servers, which then securely
forwards the message to the Client per Section 3.22.1. When the
Server forwards the NS(AR), it sets the IPv6 source address to its
own Cryptographically-Generated Address (CGA) and sets the OAL source
address to its own DLA.)
3.14.6. Route Optimization Maintenance
Following route optimization, the ROS forwards future data packets
destined to the target via the addresses found in the cached link-
layer information. The route optimization is shared by all sources
that send packets to the target via the ROS, i.e., and not just the
source on behalf of which the route optimization was initiated.
While new data packets destined to the target are flowing through the
ROS, it sends additional NS(AR) messages to the ROR before
ReachableTime expires to receive a fresh solicited NA message the
same as described in the previous sections (route optimization
refreshment strategies are an implementation matter, with a non-
normative example given in Appendix A.1). The ROS uses the cached
DLA of the ROR as the NS(AR) OAL destination address (i.e., instead
of using the DLA corresponding to the target as was the case for the
initial NS(AR)), and sends up to MAX_MULTICAST_SOLICIT NS(AR)
messages separated by 1 second until an NA is received. If no NA is
received, the ROS assumes that the current ROR has become unreachable
and deletes the target neighbor cache entry. Subsequent data packets
will trigger a new route optimization with an NS with OAL destination
address set to the DLA corresponding to the target per Section 3.14.1
to discover a new ROR while initial data packets travel over a
suboptimal route.
If an NA is received, the ROS then updates the asymmetric neighbor
cache entry to refresh ReachableTime, while (for MNP destinations)
the ROR adds or updates the ROS address to the target's Report List
and with time set to ReportTime. While no data packets are flowing,
the ROS instead allows ReachableTime for the asymmetric neighbor
cache entry to expire. When ReachableTime expires, the ROS deletes
the asymmetric neighbor cache entry. Any future data packets flowing
through the ROS will again trigger a new route optimization.
The ROS may also receive unsolicited NA messages from the ROR at any
time (see: Section 3.16). If there is an asymmetric neighbor cache
entry for the target, the ROS updates the link-layer information but
does not update ReachableTime since the receipt of an unsolicited NA
does not confirm that any forward paths are working. If there is no
asymmetric neighbor cache entry, the ROS simply discards the
unsolicited NA.
Templin Expires May 6, 2021 [Page 50]
Internet-Draft AERO November 2020
In this arrangement, the ROS holds an asymmetric neighbor cache entry
for the target via the ROR, but the ROR does not hold an asymmetric
neighbor cache entry for the ROS. The route optimization neighbor
relationship is therefore asymmetric and unidirectional. If the
target node also has packets to send back to the source node, then a
separate route optimization procedure is performed in the reverse
direction. But, there is no requirement that the forward and reverse
paths be symmetric.
3.15. Neighbor Unreachability Detection (NUD)
AERO nodes perform Neighbor Unreachability Detection (NUD) per
[RFC4861] either reactively in response to persistent link-layer
errors (see Section 3.11) or proactively to confirm reachability.
The NUD algorithm is based on periodic control message exchanges.
The algorithm may further be seeded by ND hints of forward progress,
but care must be taken to avoid inferring reachability based on
spoofed information. For example, authentic IPv6 ND message
exchanges may be considered as acceptable hints of forward progress,
while spurious data packets should not be.
AERO Servers, Proxys and Relays can use (OAL-encapsulated) standard
NS/NA NUD exchanges sent over the OMNI link spanning tree to securely
test reachability without risk of DoS attacks from nodes pretending
to be a neighbor (these NS/NA(NUD) messages use the unicast LLAs and
DLAs of the two parties involved in the NUD test the same as for
standard IPv6 ND). Proxys can further perform NUD to securely verify
Server reachability on behalf of their proxyed Clients. However, a
means for an ROS to test the unsecured forward directions of target
route optimized paths is also necessary.
When an ROR directs an ROS to a neighbor with one or more target
link-layer addresses, the ROS can proactively test each such
unsecured route optimized path by sending "loopback" NS(NUD)
messages. While testing the paths, the ROS can optionally continue
to send packets via the spanning tree, maintain a small queue of
packets until target reachability is confirmed, or (optimistically)
allow packets to flow via the route optimized paths.
When the ROS sends a loopback NS(NUD) message, it uses its own LLA as
both the IPv6 source and destination address, and the MNP Subnet-
Router anycast address as the Target Address. The ROS includes a
Nonce and Timestamp option, then encapsulates the message in OAL/INET
headers with its own DLA as the source and the DLA of the route
optimization target as the destination. The ROS then forwards the
message to the target (either directly to the L2ADDR of the target if
the target is in the same OMNI link segment, or via a Bridge if the
target is in a different OMNI link segment).
Templin Expires May 6, 2021 [Page 51]
Internet-Draft AERO November 2020
When the route optimization target receives the NS(NUD) message, it
notices that the IPv6 destination address is the same as the source
address. It then reverses the OAL header source and destination
addresses and returns the message to the ROS (either directly or via
the spanning tree). The route optimization target does not decrement
the NS(NUD) message IPv6 Hop-Limit in the process, since the message
has not exited the OMNI link.
When the ROS receives the NS(NUD) message, it can determine from the
Nonce, Timestamp and Target Address that the message originated from
itself and that it transited the forward path. The ROS need not
prepare an NA response, since the destination of the response would
be itself and testing the route optimization path again would be
redundant.
The ROS marks route optimization target paths that pass these NUD
tests as "reachable", and those that do not as "unreachable". These
markings inform the OMNI interface forwarding algorithm specified in
Section 3.10.
Note that to avoid a DoS vector nodes MUST NOT return loopback
NS(NUD) messages received from an unsecured link-layer source via the
spanning tree.
3.16. Mobility Management and Quality of Service (QoS)
AERO is a Distributed Mobility Management (DMM) service. Each Server
is responsible for only a subset of the Clients on the OMNI link, as
opposed to a Centralized Mobility Management (CMM) service where
there is a single network mobility collective entity for all Clients.
Clients coordinate with their associated Servers via RS/RA exchanges
to maintain the DMM profile, and the AERO routing system tracks all
current Client/Server peering relationships.
Servers provide default routing and mobility/multilink services for
their dependent Clients. Clients are responsible for maintaining
neighbor relationships with their Servers through periodic RS/RA
exchanges, which also serves to confirm neighbor reachability. When
a Client's underlying interface address and/or QoS information
changes, the Client is responsible for updating the Server with this
new information. Note that when there is a Proxy in the path, the
Proxy can also perform some RS/RA exchanges on the Client's behalf.
Mobility management messaging is based on the transmission and
reception of unsolicited Neighbor Advertisement (uNA) messages. Each
uNA message sets the IPv6 destination address to (link-local) All-
Nodes multicast to convey a general update of Interface Attributes to
(possibly) multiple recipients, or to a specific unicast LLA to
Templin Expires May 6, 2021 [Page 52]
Internet-Draft AERO November 2020
announce a departure event to a specific recipient. Implementations
must therefore examine the destination address to determine the
nature of the mobility event (i.e., update vs departure).
Mobility management considerations are specified in the following
sections.
3.16.1. Mobility Update Messaging
Servers accommodate Client mobility, multilink and/or QoS change
events by sending unsolicited NA (uNA) messages to each ROS in the
target Client's Report List. When a Server sends a uNA message, it
sets the IPv6 source address to the Client's LLA, sets the
destination address to (link-local) All-Nodes multicast and sets the
Target Address to the Client's Subnet-Router anycast address. The
Server also includes an OMNI option with Prefix Length set to the
length associated with the Client's LLA, with Interface Attributes
for the target Client's underlying interfaces and with the OMNI
header S/T-ifIndex set to 0. The Server then sets the NA R flag to
1, the S flag to 0 and the O flag to 1, then encapsulates the message
in an OAL header with source set to its own DLA and destination set
to the DLA of the ROS and sends the message into the spanning tree.
As discussed in Section 7.2.6 of [RFC4861], the transmission and
reception of uNA messages is unreliable but provides a useful
optimization. In well-connected Internetworks with robust data links
uNA messages will be delivered with high probability, but in any case
the Server can optionally send up to MAX_NEIGHBOR_ADVERTISEMENT uNAs
to each ROS to increase the likelihood that at least one will be
received.
When the ROS receives a uNA message prepared as above, it ignores the
message if there is no existing neighbor cache entry for the Client.
Otherwise, it uses the included OMNI option information to update the
neighbor cache entry, but does not reset ReachableTime since the
receipt of an unsolicited NA message from the target Server does not
provide confirmation that any forward paths to the target Client are
working.
If uNA messages are lost, the ROS may be left with stale address and/
or QoS information for the Client for up to ReachableTime seconds.
During this time, the ROS can continue sending packets according to
its stale neighbor cache information. When ReachableTime is close to
expiring, the ROS will re-initiate route optimization and receive
fresh link-layer address information.
In addition to sending uNA messages to the current set of ROSs for
the Client, the Server also sends uNAs to the DLA associated with the
Templin Expires May 6, 2021 [Page 53]
Internet-Draft AERO November 2020
link-layer address for any underlying interface for which the link-
layer address has changed. These uNA messages update an old Proxy/
Server that cannot easily detect (e.g., without active probing) when
a formerly-active Client has departed. When the Server sends the
uNA, it sets the IPv6 source address to the Client's LLA, sets the
destination address to the old Proxy/Server's LLA, and sets the
Target Address to the Client's Subnet-Router anycast address. The
Server also includes an OMNI option with Prefix Length set to the
length associated with the Client's LLA, with Interface Attributes
for the changed underlying interface, and with the OMNI header S/
T-ifIndex set to 0. The Server then sets the NA R flag to 1, the S
flag to 0 and the O flag to 1, then encapsulates the message in an
OAL header with source set to its own DLA and destination set to the
DLA of the old Proxy/Server and sends the message into the spanning
tree.
3.16.2. Announcing Link-Layer Address and/or QoS Preference Changes
When a Client needs to change its underlying interface addresses and/
or QoS preferences (e.g., due to a mobility event), either the Client
or its Proxys send RS messages to the Server via the spanning tree
with an OMNI option that includes Interface attributes with the new
link quality and address information.
Up to MAX_RTR_SOLICITATIONS RS messages MAY be sent in parallel with
sending actual data packets in case one or more RAs are lost. If all
RAs are lost, the Client SHOULD re-associate with a new Server.
When the Server receives the Client's changes, it sends uNA messages
to all nodes in the Report List the same as described in the previous
section.
3.16.3. Bringing New Links Into Service
When a Client needs to bring new underlying interfaces into service
(e.g., when it activates a new data link), it sends an RS message to
the Server via the underlying interface with an OMNI option that
includes Interface Attributes with appropriate link quality values
and with link-layer address information for the new link.
3.16.4. Deactivating Existing Links
When a Client needs to deactivate an existing underlying interface,
it sends an RS or uNA message to its Server with an OMNI option with
appropriate Interface Attribute values - in particular, the link
quality value 0 assures that neighbors will cease to use the link.
Templin Expires May 6, 2021 [Page 54]
Internet-Draft AERO November 2020
If the Client needs to send RS/uNA messages over an underlying
interface other than the one being deactivated, it MUST include
Interface Attributes with appropriate link quality values for any
underlying interfaces being deactivated.
Note that when a Client deactivates an underlying interface,
neighbors that have received the RS/uNA messages need not purge all
references for the underlying interface from their neighbor cache
entries. The Client may reactivate or reuse the underlying interface
and/or its ifIndex at a later point in time, when it will send RS/uNA
messages with fresh Interface Attributes to update any neighbors.
3.16.5. Moving Between Servers
The Client performs the procedures specified in Section 3.12.2 when
it first associates with a new Server or renews its association with
an existing Server. The Client also includes MS-Release identifiers
in the RS message OMNI option per [I-D.templin-6man-omni-interface]
if it wants the new Server to notify any old Servers from which the
Client is departing.
When the new Server receives the Client's RS message, it returns an
RA as specified in Section 3.12.3 and sends up to
MAX_NEIGHBOR_ADVERTIISEMENT uNA messages to any old Servers listed in
OMNI option MS-Release identifiers. When the new Server sends a uNA
message, it sets the IPv6 source address to the Client's LLA, sets
the destination address to the old Server's LLA, and sets the Target
Address to the Client's Subnet-Router anycast address. The new
Server also includes an OMNI option with Prefix Length set to the
length associated with the Client's LLA, with Interface Attributes
for its own underlying interface, and with the OMNI header S/
T-ifIndex set to 0. The new Server then sets the NA R flag to 1, the
S flag to 0 and the O flag to 1, then encapsulates the message in an
OAL header with source set to its own DLA and destination set to the
DLA of the old Server and sends the message into the spanning tree.
When an old Server receives the uNA, it changes the Client's neighbor
cache entry state to DEPARTED, sets the link-layer address of the
Client to the new Server's DLA, and resets DepartTime. After a short
delay (e.g., 2 seconds) the old Server withdraws the Client's MNP
from the routing system. After DepartTime expires, the old Server
deletes the Client's neighbor cache entry.
The old Server also iteratively forwards a copy of the uNA message to
each ROS in the Client's Report List by changing the OAL destination
address to the DLA of the ROS while leaving all other fields of the
message unmodified. When the ROS receives the uNA, it examines the
Target address to determine the correct asymmetric neighbor cache
Templin Expires May 6, 2021 [Page 55]
Internet-Draft AERO November 2020
entry and verifies that the IPv6 destination address matches the old
Server. The ROS then caches the IPv6 source address as the new
Server for the existing asymmetric neighbor cache entry and marks the
entry as STALE. While in the STALE state, the ROS allows new data
packets to flow according to any existing cached link-layer
information and sends new NS(AR) messages using its own DLA as the
OAL source and the DLA of the new Server as the OAL destination
address to elicit NA messages that reset the asymmetric neighbor
cache entry state to REACHABLE. If no new NA message is received for
10 seconds while in the STALE state, the ROS deletes the neighbor
cache entry.
Clients SHOULD NOT move rapidly between Servers in order to avoid
causing excessive oscillations in the AERO routing system. Examples
of when a Client might wish to change to a different Server include a
Server that has gone unreachable, topological movements of
significant distance, movement to a new geographic region, movement
to a new OMNI link segment, etc.
When a Client moves to a new Server, some of the fragments of a
multiple fragment packet may have already arrived at the old Server
while others are en route to the new Server, however no special
attention in the reassembly algorithm is necessary when re-routed
fragments are simply treated as loss.
3.17. Multicast
The AERO Client provides an IGMP (IPv4) [RFC2236] or MLD (IPv6)
[RFC3810] proxy service for its EUNs and/or hosted applications
[RFC4605]. The Client forwards IGMP/MLD messages over any of its
underlying interfaces for which group membership is required. The
IGMP/MLD messages may be further forwarded by a first-hop ANET access
router acting as an IGMP/MLD-snooping switch [RFC4541], then
ultimately delivered to an AERO Proxy/Server acting as a Protocol
Independent Multicast - Sparse-Mode (PIM-SM, or simply "PIM")
Designated Router (DR) [RFC7761]. AERO Relays also act as PIM
routers (i.e., the same as AERO Proxys/Servers) on behalf of nodes on
INET/EUN networks. The behaviors identified in the following
sections correspond to Source-Specific Multicast (SSM) and Any-Source
Multicast (ASM) operational modes.
3.17.1. Source-Specific Multicast (SSM)
When an ROS (i.e., an AERO Proxy/Server/Relay) "X" acting as PIM
router receives a Join/Prune message from a node on its downstream
interfaces containing one or more ((S)ource, (G)roup) pairs, it
updates its Multicast Routing Information Base (MRIB) accordingly.
For each S belonging to a prefix reachable via X's non-OMNI
Templin Expires May 6, 2021 [Page 56]
Internet-Draft AERO November 2020
interfaces, X then forwards the (S, G) Join/Prune to any PIM routers
on those interfaces per [RFC7761].
For each S belonging to a prefix reachable via X's OMNI interface, X
originates a separate copy of the Join/Prune for each (S,G) in the
message using its own LLA as the source address and ALL-PIM-ROUTERS
as the destination address. X then encapsulates each message in an
OAL header with source address set to the DLA of X and destination
address set to S then forwards the message into the spanning tree,
which delivers it to AERO Server/Relay "Y" that services S. At the
same time, if the message was a Join, X sends a route-optimization NS
message toward each S the same as discussed in Section 3.14. The
resulting NAs will return the LLA for the prefix that matches S as
the network-layer source address and with an OMNI option with the DLA
corresponding to any underlying interfaces that are currently
servicing S.
When Y processes the Join/Prune message, if S located behind any
INET, Direct, or VPNed interfaces Y acts as a PIM router and updates
its MRIB to list X as the next hop in the reverse path. If S is
located behind any Proxys "Z"*, Y also forwards the message to each
Z* over the spanning tree while continuing to use the LLA of X as the
source address. Each Z* then updates its MRIB accordingly and
maintains the LLA of X as the next hop in the reverse path. Since
the Bridges do not examine network layer control messages, this means
that the (reverse) multicast tree path is simply from each Z* (and/or
Y) to X with no other multicast-aware routers in the path. If any Z*
(and/or Y) is located on the same OMNI link segment as X, the
multicast data traffic sent to X directly using OAL/INET
encapsulation instead of via a Bridge.
Following the initial Join/Prune and NS/NA messaging, X maintains an
asymmetric neighbor cache entry for each S the same as if X was
sending unicast data traffic to S. In particular, X performs
additional NS/NA exchanges to keep the neighbor cache entry alive for
up to t_periodic seconds [RFC7761]. If no new Joins are received
within t_periodic seconds, X allows the neighbor cache entry to
expire. Finally, if X receives any additional Join/Prune messages
for (S,G) it forwards the messages to each Y and Z* in the neighbor
cache entry over the spanning tree.
At some later time, Client C that holds an MNP for source S may
depart from a first Proxy Z1 and/or connect via a new Proxy Z2. In
that case, Y sends an unsolicited NA message to X the same as
specified for unicast mobility in Section 3.16. When X receives the
unsolicited NA message, it updates its asymmetric neighbor cache
entry for the LLA for source S and sends new Join messages to any new
Proxys Z2. There is no requirement to send any Prune messages to old
Templin Expires May 6, 2021 [Page 57]
Internet-Draft AERO November 2020
Proxys Z1 since source S will no longer source any multicast data
traffic via Z1. Instead, the multicast state for (S,G) in Proxy Z1
will soon time out since no new Joins will arrive.
After some later time, C may move to a new Server Y2 and depart from
old Sever Y1. In that case, Y1 sends Join messages for any of C's
active (S,G) groups to Y2 while including its own LLA as the source
address. This causes Y2 to include Y1 in the multicast forwarding
tree during the interim time that Y1's symmetric neighbor cache entry
for C is in the DEPARTED state. At the same time, Y1 sends an
unsolicited NA message to X with an OMNI option with S/T-ifIndex in
the header set to 0 and a release indication to cause X to release
its asymmetric neighbor cache entry. X then sends a new Join message
to S via the spanning tree and re-initiates route optimization the
same as if it were receiving a fresh Join message from a node on a
downstream link.
3.17.2. Any-Source Multicast (ASM)
When an ROS X acting as a PIM router receives a Join/Prune from a
node on its downstream interfaces containing one or more (*,G) pairs,
it updates its Multicast Routing Information Base (MRIB) accordingly.
X then forwards a copy of the message to the Rendezvous Point (RP) R
for each G over the spanning tree. X uses its own LLA as the source
address and ALL-PIM-ROUTERS as the destination address, then
encapsulates each message in an OAL header with source address set to
the DLA of X and destination address set to R, then sends the message
into the spanning tree. At the same time, if the message was a Join
X initiates NS/NA route optimization the same as for the SSM case
discussed in Section 3.17.1.
For each source S that sends multicast traffic to group G via R, the
Proxy/Server Z* for the Client that aggregates S encapsulates the
packets in PIM Register messages and forwards them to R via the
spanning tree, which may then elect to send a PIM Join to Z*. This
will result in an (S,G) tree rooted at Z* with R as the next hop so
that R will begin to receive two copies of the packet; one native
copy from the (S, G) tree and a second copy from the pre-existing (*,
G) tree that still uses PIM Register encapsulation. R can then issue
a PIM Register-stop message to suppress the Register-encapsulated
stream. At some later time, if C moves to a new Proxy/Server Z*, it
resumes sending packets via PIM Register encapsulation via the new
Z*.
At the same time, as multicast listeners discover individual S's for
a given G, they can initiate an (S,G) Join for each S under the same
procedures discussed in Section 3.17.1. Once the (S,G) tree is
established, the listeners can send (S, G) Prune messages to R so
Templin Expires May 6, 2021 [Page 58]
Internet-Draft AERO November 2020
that multicast packets for group G sourced by S will only be
delivered via the (S, G) tree and not from the (*, G) tree rooted at
R. All mobility considerations discussed for SSM apply.
3.17.3. Bi-Directional PIM (BIDIR-PIM)
Bi-Directional PIM (BIDIR-PIM) [RFC5015] provides an alternate
approach to ASM that treats the Rendezvous Point (RP) as a Designated
Forwarder (DF). Further considerations for BIDIR-PIM are out of
scope.
3.18. Operation over Multiple OMNI Links
An AERO Client can connect to multiple OMNI links the same as for any
data link service. In that case, the Client maintains a distinct
OMNI interface for each link, e.g., 'omni0' for the first link,
'omni1' for the second, 'omni2' for the third, etc. Each OMNI link
would include its own distinct set of Bridges, Servers and Proxys,
thereby providing redundancy in case of failures.
Each OMNI link could utilize the same or different ANET connections.
The links can be distinguished at the link-layer via the SRT prefix
in a similar fashion as for Virtual Local Area Network (VLAN) tagging
(e.g., IEEE 802.1Q) and/or through assignment of distinct sets of
MSPs on each link. This gives rise to the opportunity for supporting
multiple redundant networked paths, with each VLAN distinguished by a
different SRT "color" (see: Section 3.2.5).
The Client's IP layer can select the outgoing OMNI interface
appropriate for a given traffic profile while (in the reverse
direction) correspondent nodes must have some way of steering their
packets destined to a target via the correct OMNI link.
In a first alternative, if each OMNI link services different MSPs,
then the Client can receive a distinct MNP from each of the links.
IP routing will therefore assure that the correct Red/Green/Blue/etc.
network is used for both outbound and inbound traffic. This can be
accomplished using existing technologies and approaches, and without
requiring any special supporting code in correspondent nodes or
Bridges.
In a second alternative, if each OMNI link services the same MSP(s)
then each link could assign a distinct "OMNI link Anycast" address
that is configured by all Bridges on the link. Correspondent nodes
can then perform Segment Routing to select the correct SRT, which
will then direct the packet over multiple hops to the target.
Templin Expires May 6, 2021 [Page 59]
Internet-Draft AERO November 2020
3.19. DNS Considerations
AERO Client MNs and INET correspondent nodes consult the Domain Name
System (DNS) the same as for any Internetworking node. When
correspondent nodes and Client MNs use different IP protocol versions
(e.g., IPv4 correspondents and IPv6 MNs), the INET DNS must maintain
A records for IPv4 address mappings to MNs which must then be
populated in Relay NAT64 mapping caches. In that way, an IPv4
correspondent node can send packets to the IPv4 address mapping of
the target MN, and the Relay will translate the IPv4 header and
destination address into an IPv6 header and IPv6 destination address
of the MN.
When an AERO Client registers with an AERO Server, the Server can
return the address(es) of DNS servers in RDNSS options [RFC6106].
The DNS server provides the IP addresses of other MNs and
correspondent nodes in AAAA records for IPv6 or A records for IPv4.
3.20. Transition Considerations
OAL encapsulation ensures that dissimilar INET partitions can be
joined into a single unified OMNI link, even though the partitions
themselves may have differing protocol versions and/or incompatible
addressing plans. However, a commonality can be achieved by
incrementally distributing globally routable (i.e., native) IP
prefixes to eventually reach all nodes (both mobile and fixed) in all
OMNI link segments. This can be accomplished by incrementally
deploying AERO Relays on each INET partition, with each Relay
distributing its MNPs and/or discovering non-MNP prefixes on its INET
links.
This gives rise to the opportunity to eventually distribute native IP
addresses to all nodes, and to present a unified OMNI link view even
if the INET partitions remain in their current protocol and
addressing plans. In that way, the OMNI link can serve the dual
purpose of providing a mobility/multilink service and a transition
service. Or, if an INET partition is transitioned to a native IP
protocol version and addressing scheme that is compatible with the
OMNI link MNP-based addressing scheme, the partition and OMNI link
can be joined by Relays.
Relays that connect INETs/EUNs with dissimilar IP protocol versions
may need to employ a network address and protocol translation
function such as NAT64 [RFC6146].
Templin Expires May 6, 2021 [Page 60]
Internet-Draft AERO November 2020
3.21. Detecting and Reacting to Server and Bridge Failures
In environments where rapid failure recovery is required, Servers and
Bridges SHOULD use Bidirectional Forwarding Detection (BFD)
[RFC5880]. Nodes that use BFD can quickly detect and react to
failures so that cached information is re-established through
alternate nodes. BFD control messaging is carried only over well-
connected ground domain networks (i.e., and not low-end radio links)
and can therefore be tuned for rapid response.
Servers and Bridges maintain BFD sessions in parallel with their BGP
peerings. If a Server or Bridge fails, BGP peers will quickly re-
establish routes through alternate paths the same as for common BGP
deployments. Similarly, Proxys maintain BFD sessions with their
associated Bridges even though they do not establish BGP peerings
with them.
Proxys SHOULD use proactive NUD for Servers for which there are
currently active ANET Clients in a manner that parallels BFD, i.e.,
by sending unicast NS messages in rapid succession to receive
solicited NA messages. When the Proxy is also sending RS messages on
behalf of ANET Clients, the RS/RA messaging can be considered as
equivalent hints of forward progress. This means that the Proxy need
not also send a periodic NS if it has already sent an RS within the
same period. If a Server fails, the Proxy will cease to receive
advertisements and can quickly inform Clients of the outage by
sending multicast RA messages on the ANET interface.
The Proxy sends multicast RA messages with source address set to the
Server's address, destination address set to (link-local) All-Nodes
multicast, and Router Lifetime set to 0. The Proxy SHOULD send
MAX_FINAL_RTR_ADVERTISEMENTS RA messages separated by small delays
[RFC4861]. Any Clients on the ANET interface that have been using
the (now defunct) Server will receive the RA messages and associate
with a new Server.
3.22. AERO Clients on the Open Internet
AERO Clients that connect to the open Internet via INET interfaces
can establish a VPN or direct link to securely connect to a Server in
a "tethered" arrangement with all of the Client's traffic transiting
the Server. Alternatively, the Client can associate with an INET
Server using UDP/IP encapsulation and asymmetric securing services as
discussed in the following sections.
When a Client's OMNI interface enables an INET underlying interface,
it first determines whether the interface is likely to be behind a
NAT. For IPv4, the Client assumes it is on the open Internet if the
Templin Expires May 6, 2021 [Page 61]
Internet-Draft AERO November 2020
INET address is not a special-use IPv4 address per [RFC3330].
Similarly for IPv6, the Client assumes it is on the open Internet if
the INET address is not a link-local [RFC4291] or unique-local
[RFC4193] IPv6 address.
The Client then prepares a UDP/IP-encapsulated RS message with IPv6
source address set to its LLA, with IPv6 destination set to (link-
local) All-Routers multicast and with an OMNI option with underlying
interface parameters. If the Client believes that it is on the open
Internet, it SHOULD include Interface Attributes with the L2ADDR used
for INET encapsulation (otherwise, it MAY omit L2ADDR). If the
underlying address is IPv4, the Client includes the Port Number and
IPv4 address written in obfuscated form [RFC4380] as discussed in
Section 3.3. If the underlying interface address is IPv6, the Client
instead includes the Port Number and IPv6 address in obfuscated form.
The Client finally includes an Authentication option per [RFC4380] to
provide message authentication, sets the UDP/IP source to its INET
address and UDP port, sets the UDP/IP destination to the Server's
INET address and the AERO service port number (8060), then sends the
message to the Server.
When the Server receives the RS, it authenticates the message and
registers the Client's MNP and INET interface information according
to the OMNI option parameters. If the RS message includes an L2ADDR
in the OMNI option, the Server compares the encapsulation IP address
and UDP port number with the (unobfuscated) values. If the values
are the same, the Server caches the Client's information as "INET"
addresses meaning that the Client is likely to accept direct messages
without requiring NAT traversal exchanges. If the values are
different (or, if the OMNI option did not include an L2ADDR) the
Server instead caches the Client's information as "NAT" addresses
meaning that NAT traversal exchanges may be necessary.
The Server then returns an RA message with IPv6 source and
destination set corresponding to the addresses in the RS, and with an
Authentication option per [RFC4380]. For IPv4, the Server also
includes an Origin option per [RFC4380] with the mapped and
obfuscated Port Number and IPv4 address observed in the encapsulation
headers. For IPv6, the Server instead includes an IPv6 Origin option
per Figure 7 with the mapped and obfuscated observed Port Number and
IPv6 address (note that the value 0x02 in the second octet
differentiates from other [RFC4380] option types).
Templin Expires May 6, 2021 [Page 62]
Internet-Draft AERO November 2020
+--------+--------+-----------------+
| 0x00 | 0x02 | Origin port # |
+--------+--------+-----------------+
~ Origin IPv6 address ~
+-----------------------------------+
Figure 7: IPv6 Origin Option
When the Client receives the RA message, it compares the mapped Port
Number and IP address from the Origin option with its own address.
If the addresses are the same, the Client assumes the open Internet /
Cone NAT principle; if the addresses are different, the Client
instead assumes that further qualification procedures are necessary
to detect the type of NAT and proceeds according to standard
[RFC4380] procedures.
After the Client has registered its INET interfaces in such RS/RA
exchanges it sends periodic RS messages to receive fresh RA messages
before the Router Lifetime received on each INET interface expires.
The Client also maintains default routes via its Servers, i.e., the
same as described in earlier sections.
When the Client sends messages to target IP addresses, it also
invokes route optimization per Section 3.14 using IPv6 ND address
resolution messaging. The Client sends the NS(AR) message to the
Server wrapped in a UDP/IP header with an Authentication option with
the NS source address set to the Client's LLA and destination address
set to the target solicited node multicast address. The Server
authenticates the message and sends a corresponding NS(AR) message
over the spanning tree the same as if it were the ROS, but with the
OAL source address set to the Server's DLA and destination set to the
DLA of the target. When the ROR receives the NS(AR), it adds the
Server's DLA and Client's LLA to the target's Report List, and
returns an NA with OMNI option information for the target. The
Server then returns a UDP/IP encapsulated NA message with an
Authentication option to the Client.
Following route optimization for targets in the same OMNI link
segment, if the target's L2ADDR is on the open INET, the Client
forwards data packets directly to the target INET address. If the
target is behind a NAT, the Client first establishes NAT state for
the L2ADDR using the "bubble" mechanisms specified in
[RFC6081][RFC4380]. The Client continues to send data packets via
its Server until NAT state is populated, then begins forwarding
packets via the direct path through the NAT to the target. For
targets in different OMNI link segments, the Client inserts an ORH
and forwards data packets to the Bridge that returned the NA message.
Templin Expires May 6, 2021 [Page 63]
Internet-Draft AERO November 2020
The ROR may return uNAs via the Server if the target moves, and the
Server will send corresponding Authentication-protected uNAs to the
Client. The Client can also send "loopback" NS(NUD) messages to test
forward path reachability even though there is no security
association between the Client and the target.
The Client sends UDP/IP encapsulated IPv6 packets no larger than 1280
bytes in one piece. In order to accommodate larger IPv6 packets (up
to the OMNI interface MTU), the Client inserts an OAL header with
source set to its own DLA and destination set to the DLA of the
target and uses IPv6 fragmentation according to Section 3.9. The
Client then encapsulates each fragment in a UDP/IP header and sends
the fragments to the next hop.
3.22.1. Use of SEND and CGA
In some environments, use of the [RFC4380] Authentication option
alone may be sufficient for assuring IPv6 ND message authentication
between Clients and Servers. When additional protection is
necessary, nodes should employ SEcure Neighbor Discovery (SEND)
[RFC3971] with Cryptographically-Generated Addresses (CGA) [RFC3972].
When SEND/CGA are used, the Client prepares RS messages with its
link-local CGA as the IPv6 source and (link-local) All-Routers
multicast as the IPv6 Destination, includes any SEND options and
wraps the message in an OAL header. The Client sets the OAL source
address to its own DLA and sets the OAL destination address to (site-
local) All-Routers multicast. The Client then wraps the RS message
in UDP/IP headers according to [RFC4380] and sends the message to the
Server.
When the Server receives the message, it first verifies the
Authentication option (if present) then uses the OAL source address
to determine the MNP of the Client. The Server then processes the
SEND options to authenticate the RS message and prepares an RA
message response. The Server prepares the RA with its own link-local
CGA as the IPv6 source and the CGA of the Client as the IPv6
destination, includes any SEND options and wraps the message in an
OAL header. The Server sets the OAL source address to its own DLA
and sets the OAL destination address to the Client's DLA. The Server
then wraps the RA message in UDP/IP headers according to [RFC4380]
and sends the message to the Client. Thereafter, the Client/Server
send additional RS/RA messages to maintain their association and any
NAT state.
The Client and Server also may exchange NS/NA messages using their
own CGA as the source and with OAL encapsulation as above. When a
Client sends an NS(AR), it sets the IPv6 source to its CGA and sets
Templin Expires May 6, 2021 [Page 64]
Internet-Draft AERO November 2020
the IPv6 destination to the Solicited-Node Multicast address of the
target. The Client then wraps the message in an OAL header with its
own DLA as the source and the DLA of the target as the destination
and sends it to the Server. The Server authenticates the message,
then changes the IPv6 source address to the Client's LLA, removes the
SEND options, and sends a corresponding NS(AR) into the spanning
tree. When the Server receives the corresponding OAL-encapsulated
NA, it changes the IPv6 destination address to the Client's CGA,
inserts SEND options, then wraps the message in UDP/IP headers and
sends it to the Client.
When a Client sends a uNA, it sets the IPv6 source address to its own
CGA and sets the IPv6 destination address to (link-local) All-Nodes
multicast, includes SEND options, wraps the message in OAL and UDP/IP
headers and sends the message to the Server. The Server
authenticates the message, then changes the IPv6 address to the
Client's LLA, removes the SEND options and forwards the message the
same as discussed in Section 3.16.1. In the reverse direction, when
the Server forwards a uNA to the Client, it changes the IPv6 address
to its own CGA and inserts SEND options then forwards the message to
the Client.
When a Client sends an NS(NUD), it sets both the IPv6 source and
destination address to its own LLA, wraps the message in an OAL
header and UDP/IP headers, then sends the message directly to the
peer which will loop the message back. In this case alone, the
Client does not use the Server as a trust broker for forwarding the
ND message.
3.23. Time-Varying MNPs
In some use cases, it is desirable, beneficial and efficient for the
Client to receive a constant MNP that travels with the Client
wherever it moves. For example, this would allow air traffic
controllers to easily track aircraft, etc. In other cases, however
(e.g., intelligent transportation systems), the MN may be willing to
sacrifice a modicum of efficiency in order to have time-varying MNPs
that can be changed every so often to defeat adversarial tracking.
The DHCPv6 service offers a way for Clients that desire time-varying
MNPs to obtain short-lived prefixes (e.g., on the order of a small
number of minutes). In that case, the identity of the Client would
not be bound to the MNP but rather the Client's identity would be
bound to the DHCPv6 Device Unique Identifier (DUID) and used as the
seed for Prefix Delegation. The Client would then be obligated to
renumber its internal networks whenever its MNP (and therefore also
its LLA) changes. This should not present a challenge for Clients
with automated network renumbering services, however presents limits
Templin Expires May 6, 2021 [Page 65]
Internet-Draft AERO November 2020
for the durations of ongoing sessions that would prefer to use a
constant address.
4. Implementation Status
An early AERO implementation based on OpenVPN (https://openvpn.net/)
was announced on the v6ops mailing list on January 10, 2018 and an
initial public release of the AERO proof-of-concept source code was
announced on the intarea mailing list on August 21, 2015.
AERO Release-3.0.2 was tagged on October 15, 2020, and is undergoing
internal testing. Additional releases expected Q42020, with first
public release expected before year-end.
5. IANA Considerations
The IANA has assigned a 4-octet Private Enterprise Number "45282" for
AERO in the "enterprise-numbers" registry.
The IANA has assigned the UDP port number "8060" for an earlier
experimental version of AERO [RFC6706]. This document obsoletes
[RFC6706] and claims the UDP port number "8060" for all future use.
The IANA is instructed to assign a new type value TBD in the IPv6
Routing Types registry.
No further IANA actions are required.
6. Security Considerations
AERO Bridges configure secured tunnels with AERO Servers, Relays and
Proxys within their local OMNI link segments. Applicable secured
tunnel alternatives include IPsec [RFC4301], TLS/SSL [RFC8446], DTLS
[RFC6347], WireGuard [WG], etc. The AERO Bridges of all OMNI link
segments in turn configure secured tunnels for their neighboring AERO
Bridges in a spanning tree topology. Therefore, control messages
exchanged between any pair of OMNI link neighbors on the spanning
tree are already secured.
AERO Servers, Relays and Proxys targeted by a route optimization may
also receive data packets directly from arbitrary nodes in INET
partitions instead of via the spanning tree. For INET partitions
that apply effective ingress filtering to defeat source address
spoofing, the simple data origin authentication procedures in
Section 3.8 can be applied.
For INET partitions that require strong security in the data plane,
two options for securing communications include 1) disable route
Templin Expires May 6, 2021 [Page 66]
Internet-Draft AERO November 2020
optimization so that all traffic is conveyed over secured tunnels, or
2) enable on-demand secure tunnel creation between INET partition
neighbors. Option 1) would result in longer routes than necessary
and traffic concentration on critical infrastructure elements.
Option 2) could be coordinated by establishing a secured tunnel on-
demand instead of performing an NS/NA exchange in the route
optimization procedures. Procedures for establishing on-demand
secured tunnels are out of scope.
AERO Clients that connect to secured ANETs need not apply security to
their ND messages, since the messages will be intercepted by a
perimeter Proxy that applies security on its INET-facing interface as
part of the spanning tree (see above). AERO Clients connected to the
open INET can use symmetric network and/or transport layer security
services such as VPNs or can by some other means establish a direct
link. When a VPN or direct link may be impractical, however, an
asymmetric security service such as SEcure Neighbor Discovery (SEND)
[RFC3971] with Cryptographically Generated Addresses (CGAs) [RFC3972]
and/or the Authentication option [RFC4380] can be applied.
Application endpoints SHOULD use application-layer security services
such as TLS/SSL, DTLS or SSH [RFC4251] to assure the same level of
protection as for critical secured Internet services. AERO Clients
that require host-based VPN services SHOULD use symmetric network
and/or transport layer security services such as IPsec, TLS/SSL,
DTLS, etc. AERO Proxys and Servers can also provide a network-based
VPN service on behalf of the Client, e.g., if the Client is located
within a secured enclave and cannot establish a VPN on its own
behalf.
AERO Servers and Bridges present targets for traffic amplification
Denial of Service (DoS) attacks. This concern is no different than
for widely-deployed VPN security gateways in the Internet, where
attackers could send spoofed packets to the gateways at high data
rates. This can be mitigated by connecting Servers and Bridges over
dedicated links with no connections to the Internet and/or when
connections to the Internet are only permitted through well-managed
firewalls. Traffic amplification DoS attacks can also target an AERO
Client's low data rate links. This is a concern not only for Clients
located on the open Internet but also for Clients in secured
enclaves. AERO Servers and Proxys can institute rate limits that
protect Clients from receiving packet floods that could DoS low data
rate links.
AERO Relays must implement ingress filtering to avoid a spoofing
attack in which spurious messages with DLA addresses are injected
into an OMNI link from an outside attacker. AERO Clients MUST ensure
that their connectivity is not used by unauthorized nodes on their
Templin Expires May 6, 2021 [Page 67]
Internet-Draft AERO November 2020
EUNs to gain access to a protected network, i.e., AERO Clients that
act as routers MUST NOT provide routing services for unauthorized
nodes. (This concern is no different than for ordinary hosts that
receive an IP address delegation but then "share" the address with
other nodes via some form of Internet connection sharing such as
tethering.)
The MAP list MUST be well-managed and secured from unauthorized
tampering, even though the list contains only public information.
The MAP list can be conveyed to the Client in a similar fashion as in
[RFC5214] (e.g., through layer 2 data link login messaging, secure
upload of a static file, DNS lookups, etc.).
Although public domain and commercial SEND implementations exist,
concerns regarding the strength of the cryptographic hash algorithm
have been documented [RFC6273] [RFC4982].
SRH authentication facilities are specified in [RFC8754].
Security considerations for accepting link-layer ICMP messages and
reflected packets are discussed throughout the document.
Security considerations for IPv6 fragmentation and reassembly are
discussed in [I-D.templin-6man-omni-interface].
7. Acknowledgements
Discussions in the IETF, aviation standards communities and private
exchanges helped shape some of the concepts in this work.
Individuals who contributed insights include Mikael Abrahamsson, Mark
Andrews, Fred Baker, Bob Braden, Stewart Bryant, Brian Carpenter,
Wojciech Dec, Pavel Drasil, Ralph Droms, Adrian Farrel, Nick Green,
Sri Gundavelli, Brian Haberman, Bernhard Haindl, Joel Halpern, Tom
Herbert, Sascha Hlusiak, Lee Howard, Zdenek Jaron, Andre Kostur,
Hubert Kuenig, Ted Lemon, Andy Malis, Satoru Matsushima, Tomek
Mrugalski, Madhu Niraula, Alexandru Petrescu, Behcet Saikaya, Michal
Skorepa, Joe Touch, Bernie Volz, Ryuji Wakikawa, Tony Whyman, Lloyd
Wood and James Woodyatt. Members of the IESG also provided valuable
input during their review process that greatly improved the document.
Special thanks go to Stewart Bryant, Joel Halpern and Brian Haberman
for their shepherding guidance during the publication of the AERO
first edition.
This work has further been encouraged and supported by Boeing
colleagues including Kyle Bae, M. Wayne Benson, Dave Bernhardt, Cam
Brodie, John Bush, Balaguruna Chidambaram, Irene Chin, Bruce Cornish,
Claudiu Danilov, Don Dillenburg, Joe Dudkowski, Wen Fang, Samad
Farooqui, Anthony Gregory, Jeff Holland, Seth Jahne, Brian Jaury,
Templin Expires May 6, 2021 [Page 68]
Internet-Draft AERO November 2020
Greg Kimberly, Ed King, Madhuri Madhava Badgandi, Laurel Matthew,
Gene MacLean III, Rob Muszkiewicz, Sean O'Sullivan, Vijay
Rajagopalan, Greg Saccone, Rod Santiago, Kent Shuey, Brian Skeen,
Mike Slane, Carrie Spiker, Katie Tran, Brendan Williams, Amelia
Wilson, Julie Wulff, Yueli Yang, Eric Yeh and other members of the
Boeing mobility, networking and autonomy teams. Kyle Bae, Wayne
Benson, Katie Tran and Eric Yeh are especially acknowledged for
implementing the AERO functions as extensions to the public domain
OpenVPN distribution.
Earlier works on NBMA tunneling approaches are found in
[RFC2529][RFC5214][RFC5569].
Many of the constructs presented in this second edition of AERO are
based on the author's earlier works, including:
o The Internet Routing Overlay Network (IRON)
[RFC6179][I-D.templin-ironbis]
o Virtual Enterprise Traversal (VET)
[RFC5558][I-D.templin-intarea-vet]
o The Subnetwork Encapsulation and Adaptation Layer (SEAL)
[RFC5320][I-D.templin-intarea-seal]
o AERO, First Edition [RFC6706]
Note that these works cite numerous earlier efforts that are not also
cited here due to space limitations. The authors of those earlier
works are acknowledged for their insights.
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 Commercial Airplanes (BCA)
Internet of Things (IoT) and autonomy programs.
This work is aligned with the Boeing Information Technology (BIT)
MobileNet program.
8. References
Templin Expires May 6, 2021 [Page 69]
Internet-Draft AERO November 2020
8.1. Normative References
[I-D.templin-6man-omni-interface]
Templin, F. and T. Whyman, "Transmission of IP Packets
over Overlay Multilink Network (OMNI) Interfaces", draft-
templin-6man-omni-interface-50 (work in progress), October
2020.
[RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791,
DOI 10.17487/RFC0791, September 1981,
<https://www.rfc-editor.org/info/rfc791>.
[RFC0792] Postel, J., "Internet Control Message Protocol", STD 5,
RFC 792, DOI 10.17487/RFC0792, September 1981,
<https://www.rfc-editor.org/info/rfc792>.
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119,
DOI 10.17487/RFC2119, March 1997,
<https://www.rfc-editor.org/info/rfc2119>.
[RFC2473] Conta, A. and S. Deering, "Generic Packet Tunneling in
IPv6 Specification", RFC 2473, DOI 10.17487/RFC2473,
December 1998, <https://www.rfc-editor.org/info/rfc2473>.
[RFC3971] Arkko, J., Ed., Kempf, J., Zill, B., and P. Nikander,
"SEcure Neighbor Discovery (SEND)", RFC 3971,
DOI 10.17487/RFC3971, March 2005,
<https://www.rfc-editor.org/info/rfc3971>.
[RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)",
RFC 3972, DOI 10.17487/RFC3972, March 2005,
<https://www.rfc-editor.org/info/rfc3972>.
[RFC4191] Draves, R. and D. Thaler, "Default Router Preferences and
More-Specific Routes", RFC 4191, DOI 10.17487/RFC4191,
November 2005, <https://www.rfc-editor.org/info/rfc4191>.
[RFC4193] Hinden, R. and B. Haberman, "Unique Local IPv6 Unicast
Addresses", RFC 4193, DOI 10.17487/RFC4193, October 2005,
<https://www.rfc-editor.org/info/rfc4193>.
[RFC4380] Huitema, C., "Teredo: Tunneling IPv6 over UDP through
Network Address Translations (NATs)", RFC 4380,
DOI 10.17487/RFC4380, February 2006,
<https://www.rfc-editor.org/info/rfc4380>.
Templin Expires May 6, 2021 [Page 70]
Internet-Draft AERO November 2020
[RFC4861] Narten, T., Nordmark, E., Simpson, W., and H. Soliman,
"Neighbor Discovery for IP version 6 (IPv6)", RFC 4861,
DOI 10.17487/RFC4861, September 2007,
<https://www.rfc-editor.org/info/rfc4861>.
[RFC4862] Thomson, S., Narten, T., and T. Jinmei, "IPv6 Stateless
Address Autoconfiguration", RFC 4862,
DOI 10.17487/RFC4862, September 2007,
<https://www.rfc-editor.org/info/rfc4862>.
[RFC6081] Thaler, D., "Teredo Extensions", RFC 6081,
DOI 10.17487/RFC6081, January 2011,
<https://www.rfc-editor.org/info/rfc6081>.
[RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
May 2017, <https://www.rfc-editor.org/info/rfc8174>.
[RFC8200] Deering, S. and R. Hinden, "Internet Protocol, Version 6
(IPv6) Specification", STD 86, RFC 8200,
DOI 10.17487/RFC8200, July 2017,
<https://www.rfc-editor.org/info/rfc8200>.
[RFC8415] Mrugalski, T., Siodelski, M., Volz, B., Yourtchenko, A.,
Richardson, M., Jiang, S., Lemon, T., and T. Winters,
"Dynamic Host Configuration Protocol for IPv6 (DHCPv6)",
RFC 8415, DOI 10.17487/RFC8415, November 2018,
<https://www.rfc-editor.org/info/rfc8415>.
8.2. Informative References
[BGP] Huston, G., "BGP in 2015, http://potaroo.net", January
2016.
[I-D.bonica-6man-comp-rtg-hdr]
Bonica, R., Kamite, Y., Alston, A., Henriques, D., and L.
Jalil, "The IPv6 Compact Routing Header (CRH)", draft-
bonica-6man-comp-rtg-hdr-23 (work in progress), October
2020.
[I-D.bonica-6man-crh-helper-opt]
Li, X., Bao, C., Ruan, E., and R. Bonica, "Compressed
Routing Header (CRH) Helper Option", draft-bonica-6man-
crh-helper-opt-02 (work in progress), October 2020.
Templin Expires May 6, 2021 [Page 71]
Internet-Draft AERO November 2020
[I-D.ietf-intarea-frag-fragile]
Bonica, R., Baker, F., Huston, G., Hinden, R., Troan, O.,
and F. Gont, "IP Fragmentation Considered Fragile", draft-
ietf-intarea-frag-fragile-17 (work in progress), September
2019.
[I-D.ietf-intarea-tunnels]
Touch, J. and M. Townsley, "IP Tunnels in the Internet
Architecture", draft-ietf-intarea-tunnels-10 (work in
progress), September 2019.
[I-D.ietf-rtgwg-atn-bgp]
Templin, F., Saccone, G., Dawra, G., Lindem, A., and V.
Moreno, "A Simple BGP-based Mobile Routing System for the
Aeronautical Telecommunications Network", draft-ietf-
rtgwg-atn-bgp-06 (work in progress), June 2020.
[I-D.templin-6man-dhcpv6-ndopt]
Templin, F., "A Unified Stateful/Stateless Configuration
Service for IPv6", draft-templin-6man-dhcpv6-ndopt-10
(work in progress), June 2020.
[I-D.templin-intarea-seal]
Templin, F., "The Subnetwork Encapsulation and Adaptation
Layer (SEAL)", draft-templin-intarea-seal-68 (work in
progress), January 2014.
[I-D.templin-intarea-vet]
Templin, F., "Virtual Enterprise Traversal (VET)", draft-
templin-intarea-vet-40 (work in progress), May 2013.
[I-D.templin-ironbis]
Templin, F., "The Interior Routing Overlay Network
(IRON)", draft-templin-ironbis-16 (work in progress),
March 2014.
[I-D.templin-v6ops-pdhost]
Templin, F., "IPv6 Prefix Delegation and Multi-Addressing
Models", draft-templin-v6ops-pdhost-26 (work in progress),
June 2020.
[OVPN] OpenVPN, O., "http://openvpn.net", October 2016.
[RFC1035] Mockapetris, P., "Domain names - implementation and
specification", STD 13, RFC 1035, DOI 10.17487/RFC1035,
November 1987, <https://www.rfc-editor.org/info/rfc1035>.
Templin Expires May 6, 2021 [Page 72]
Internet-Draft AERO November 2020
[RFC1812] Baker, F., Ed., "Requirements for IP Version 4 Routers",
RFC 1812, DOI 10.17487/RFC1812, June 1995,
<https://www.rfc-editor.org/info/rfc1812>.
[RFC2003] Perkins, C., "IP Encapsulation within IP", RFC 2003,
DOI 10.17487/RFC2003, October 1996,
<https://www.rfc-editor.org/info/rfc2003>.
[RFC2236] Fenner, W., "Internet Group Management Protocol, Version
2", RFC 2236, DOI 10.17487/RFC2236, November 1997,
<https://www.rfc-editor.org/info/rfc2236>.
[RFC2464] Crawford, M., "Transmission of IPv6 Packets over Ethernet
Networks", RFC 2464, DOI 10.17487/RFC2464, December 1998,
<https://www.rfc-editor.org/info/rfc2464>.
[RFC2529] Carpenter, B. and C. Jung, "Transmission of IPv6 over IPv4
Domains without Explicit Tunnels", RFC 2529,
DOI 10.17487/RFC2529, March 1999,
<https://www.rfc-editor.org/info/rfc2529>.
[RFC2983] Black, D., "Differentiated Services and Tunnels",
RFC 2983, DOI 10.17487/RFC2983, October 2000,
<https://www.rfc-editor.org/info/rfc2983>.
[RFC3168] Ramakrishnan, K., Floyd, S., and D. Black, "The Addition
of Explicit Congestion Notification (ECN) to IP",
RFC 3168, DOI 10.17487/RFC3168, September 2001,
<https://www.rfc-editor.org/info/rfc3168>.
[RFC3330] IANA, "Special-Use IPv4 Addresses", RFC 3330,
DOI 10.17487/RFC3330, September 2002,
<https://www.rfc-editor.org/info/rfc3330>.
[RFC3810] Vida, R., Ed. and L. Costa, Ed., "Multicast Listener
Discovery Version 2 (MLDv2) for IPv6", RFC 3810,
DOI 10.17487/RFC3810, June 2004,
<https://www.rfc-editor.org/info/rfc3810>.
[RFC4251] Ylonen, T. and C. Lonvick, Ed., "The Secure Shell (SSH)
Protocol Architecture", RFC 4251, DOI 10.17487/RFC4251,
January 2006, <https://www.rfc-editor.org/info/rfc4251>.
[RFC4271] Rekhter, Y., Ed., Li, T., Ed., and S. Hares, Ed., "A
Border Gateway Protocol 4 (BGP-4)", RFC 4271,
DOI 10.17487/RFC4271, January 2006,
<https://www.rfc-editor.org/info/rfc4271>.
Templin Expires May 6, 2021 [Page 73]
Internet-Draft AERO November 2020
[RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing
Architecture", RFC 4291, DOI 10.17487/RFC4291, February
2006, <https://www.rfc-editor.org/info/rfc4291>.
[RFC4301] Kent, S. and K. Seo, "Security Architecture for the
Internet Protocol", RFC 4301, DOI 10.17487/RFC4301,
December 2005, <https://www.rfc-editor.org/info/rfc4301>.
[RFC4389] Thaler, D., Talwar, M., and C. Patel, "Neighbor Discovery
Proxies (ND Proxy)", RFC 4389, DOI 10.17487/RFC4389, April
2006, <https://www.rfc-editor.org/info/rfc4389>.
[RFC4443] Conta, A., Deering, S., and M. Gupta, Ed., "Internet
Control Message Protocol (ICMPv6) for the Internet
Protocol Version 6 (IPv6) Specification", STD 89,
RFC 4443, DOI 10.17487/RFC4443, March 2006,
<https://www.rfc-editor.org/info/rfc4443>.
[RFC4511] Sermersheim, J., Ed., "Lightweight Directory Access
Protocol (LDAP): The Protocol", RFC 4511,
DOI 10.17487/RFC4511, June 2006,
<https://www.rfc-editor.org/info/rfc4511>.
[RFC4541] Christensen, M., Kimball, K., and F. Solensky,
"Considerations for Internet Group Management Protocol
(IGMP) and Multicast Listener Discovery (MLD) Snooping
Switches", RFC 4541, DOI 10.17487/RFC4541, May 2006,
<https://www.rfc-editor.org/info/rfc4541>.
[RFC4605] Fenner, B., He, H., Haberman, B., and H. Sandick,
"Internet Group Management Protocol (IGMP) / Multicast
Listener Discovery (MLD)-Based Multicast Forwarding
("IGMP/MLD Proxying")", RFC 4605, DOI 10.17487/RFC4605,
August 2006, <https://www.rfc-editor.org/info/rfc4605>.
[RFC4982] Bagnulo, M. and J. Arkko, "Support for Multiple Hash
Algorithms in Cryptographically Generated Addresses
(CGAs)", RFC 4982, DOI 10.17487/RFC4982, July 2007,
<https://www.rfc-editor.org/info/rfc4982>.
[RFC5015] Handley, M., Kouvelas, I., Speakman, T., and L. Vicisano,
"Bidirectional Protocol Independent Multicast (BIDIR-
PIM)", RFC 5015, DOI 10.17487/RFC5015, October 2007,
<https://www.rfc-editor.org/info/rfc5015>.
Templin Expires May 6, 2021 [Page 74]
Internet-Draft AERO November 2020
[RFC5214] Templin, F., Gleeson, T., and D. Thaler, "Intra-Site
Automatic Tunnel Addressing Protocol (ISATAP)", RFC 5214,
DOI 10.17487/RFC5214, March 2008,
<https://www.rfc-editor.org/info/rfc5214>.
[RFC5320] Templin, F., Ed., "The Subnetwork Encapsulation and
Adaptation Layer (SEAL)", RFC 5320, DOI 10.17487/RFC5320,
February 2010, <https://www.rfc-editor.org/info/rfc5320>.
[RFC5522] Eddy, W., Ivancic, W., and T. Davis, "Network Mobility
Route Optimization Requirements for Operational Use in
Aeronautics and Space Exploration Mobile Networks",
RFC 5522, DOI 10.17487/RFC5522, October 2009,
<https://www.rfc-editor.org/info/rfc5522>.
[RFC5558] Templin, F., Ed., "Virtual Enterprise Traversal (VET)",
RFC 5558, DOI 10.17487/RFC5558, February 2010,
<https://www.rfc-editor.org/info/rfc5558>.
[RFC5569] Despres, R., "IPv6 Rapid Deployment on IPv4
Infrastructures (6rd)", RFC 5569, DOI 10.17487/RFC5569,
January 2010, <https://www.rfc-editor.org/info/rfc5569>.
[RFC5880] Katz, D. and D. Ward, "Bidirectional Forwarding Detection
(BFD)", RFC 5880, DOI 10.17487/RFC5880, June 2010,
<https://www.rfc-editor.org/info/rfc5880>.
[RFC6106] Jeong, J., Park, S., Beloeil, L., and S. Madanapalli,
"IPv6 Router Advertisement Options for DNS Configuration",
RFC 6106, DOI 10.17487/RFC6106, November 2010,
<https://www.rfc-editor.org/info/rfc6106>.
[RFC6146] Bagnulo, M., Matthews, P., and I. van Beijnum, "Stateful
NAT64: Network Address and Protocol Translation from IPv6
Clients to IPv4 Servers", RFC 6146, DOI 10.17487/RFC6146,
April 2011, <https://www.rfc-editor.org/info/rfc6146>.
[RFC6179] Templin, F., Ed., "The Internet Routing Overlay Network
(IRON)", RFC 6179, DOI 10.17487/RFC6179, March 2011,
<https://www.rfc-editor.org/info/rfc6179>.
[RFC6221] Miles, D., Ed., Ooghe, S., Dec, W., Krishnan, S., and A.
Kavanagh, "Lightweight DHCPv6 Relay Agent", RFC 6221,
DOI 10.17487/RFC6221, May 2011,
<https://www.rfc-editor.org/info/rfc6221>.
Templin Expires May 6, 2021 [Page 75]
Internet-Draft AERO November 2020
[RFC6273] Kukec, A., Krishnan, S., and S. Jiang, "The Secure
Neighbor Discovery (SEND) Hash Threat Analysis", RFC 6273,
DOI 10.17487/RFC6273, June 2011,
<https://www.rfc-editor.org/info/rfc6273>.
[RFC6347] Rescorla, E. and N. Modadugu, "Datagram Transport Layer
Security Version 1.2", RFC 6347, DOI 10.17487/RFC6347,
January 2012, <https://www.rfc-editor.org/info/rfc6347>.
[RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label
for Equal Cost Multipath Routing and Link Aggregation in
Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011,
<https://www.rfc-editor.org/info/rfc6438>.
[RFC6706] Templin, F., Ed., "Asymmetric Extended Route Optimization
(AERO)", RFC 6706, DOI 10.17487/RFC6706, August 2012,
<https://www.rfc-editor.org/info/rfc6706>.
[RFC6935] Eubanks, M., Chimento, P., and M. Westerlund, "IPv6 and
UDP Checksums for Tunneled Packets", RFC 6935,
DOI 10.17487/RFC6935, April 2013,
<https://www.rfc-editor.org/info/rfc6935>.
[RFC6936] Fairhurst, G. and M. Westerlund, "Applicability Statement
for the Use of IPv6 UDP Datagrams with Zero Checksums",
RFC 6936, DOI 10.17487/RFC6936, April 2013,
<https://www.rfc-editor.org/info/rfc6936>.
[RFC7333] Chan, H., Ed., Liu, D., Seite, P., Yokota, H., and J.
Korhonen, "Requirements for Distributed Mobility
Management", RFC 7333, DOI 10.17487/RFC7333, August 2014,
<https://www.rfc-editor.org/info/rfc7333>.
[RFC7761] Fenner, B., Handley, M., Holbrook, H., Kouvelas, I.,
Parekh, R., Zhang, Z., and L. Zheng, "Protocol Independent
Multicast - Sparse Mode (PIM-SM): Protocol Specification
(Revised)", STD 83, RFC 7761, DOI 10.17487/RFC7761, March
2016, <https://www.rfc-editor.org/info/rfc7761>.
[RFC8402] Filsfils, C., Ed., Previdi, S., Ed., Ginsberg, L.,
Decraene, B., Litkowski, S., and R. Shakir, "Segment
Routing Architecture", RFC 8402, DOI 10.17487/RFC8402,
July 2018, <https://www.rfc-editor.org/info/rfc8402>.
[RFC8446] Rescorla, E., "The Transport Layer Security (TLS) Protocol
Version 1.3", RFC 8446, DOI 10.17487/RFC8446, August 2018,
<https://www.rfc-editor.org/info/rfc8446>.
Templin Expires May 6, 2021 [Page 76]
Internet-Draft AERO November 2020
[RFC8754] Filsfils, C., Ed., Dukes, D., Ed., Previdi, S., Leddy, J.,
Matsushima, S., and D. Voyer, "IPv6 Segment Routing Header
(SRH)", RFC 8754, DOI 10.17487/RFC8754, March 2020,
<https://www.rfc-editor.org/info/rfc8754>.
[WG] Wireguard, "Wireguard, https://www.wireguard.com", August
2020.
Appendix A. Non-Normative Considerations
AERO can be applied to a multitude of Internetworking scenarios, with
each having its own adaptations. The following considerations are
provided as non-normative guidance:
A.1. Implementation Strategies for Route Optimization
Route optimization as discussed in Section 3.14 results in the route
optimization source (ROS) creating an asymmetric neighbor cache entry
for the target neighbor. The neighbor cache entry is maintained for
at most ReachableTime seconds and then deleted unless updated. In
order to refresh the neighbor cache entry lifetime before the
ReachableTime timer expires, the specification requires
implementations to issue a new NS/NA exchange to reset ReachableTime
while data packets are still flowing. However, the decision of when
to initiate a new NS/NA exchange and to perpetuate the process is
left as an implementation detail.
One possible strategy may be to monitor the neighbor cache entry
watching for data packets for (ReachableTime - 5) seconds. If any
data packets have been sent to the neighbor within this timeframe,
then send an NS to receive a new NA. If no data packets have been
sent, wait for 5 additional seconds and send an immediate NS if any
data packets are sent within this "expiration pending" 5 second
window. If no additional data packets are sent within the 5 second
window, delete the neighbor cache entry.
The monitoring of the neighbor data packet traffic therefore becomes
an asymmetric ongoing process during the neighbor cache entry
lifetime. If the neighbor cache entry expires, future data packets
will trigger a new NS/NA exchange while the packets themselves are
delivered over a longer path until route optimization state is re-
established.
A.2. Implicit Mobility Management
OMNI interface neighbors MAY provide a configuration option that
allows them to perform implicit mobility management in which no ND
messaging is used. In that case, the Client only transmits packets
Templin Expires May 6, 2021 [Page 77]
Internet-Draft AERO November 2020
over a single interface at a time, and the neighbor always observes
packets arriving from the Client from the same link-layer source
address.
If the Client's underlying interface address changes (either due to a
readdressing of the original interface or switching to a new
interface) the neighbor immediately updates the neighbor cache entry
for the Client and begins accepting and sending packets according to
the Client's new address. This implicit mobility method applies to
use cases such as cellphones with both WiFi and Cellular interfaces
where only one of the interfaces is active at a given time, and the
Client automatically switches over to the backup interface if the
primary interface fails.
A.3. Direct Underlying Interfaces
When a Client's OMNI interface is configured over a Direct interface,
the neighbor at the other end of the Direct link can receive packets
without any encapsulation. In that case, the Client sends packets
over the Direct link according to QoS preferences. If the Direct
interface has the highest QoS preference, then the Client's IP
packets are transmitted directly to the peer without going through an
ANET/INET. If other interfaces have higher QoS preferences, then the
Client's IP packets are transmitted via a different interface, which
may result in the inclusion of Proxys, Servers and Bridges in the
communications path. Direct interfaces must be tested periodically
for reachability, e.g., via NUD.
A.4. AERO Critical Infrastructure Considerations
AERO Bridges can be either Commercial off-the Shelf (COTS) standard
IP routers or virtual machines in the cloud. Bridges must be
provisioned, supported and managed by the INET administrative
authority, and connected to the Bridges of other INETs via inter-
domain peerings. Cost for purchasing, configuring and managing
Bridges is nominal even for very large OMNI links.
AERO Servers can be standard dedicated server platforms, but most
often will be deployed as virtual machines in the cloud. The only
requirements for Servers are that they can run the AERO user-level
code and have at least one network interface connection to the INET.
As with Bridges, Servers must be provisioned, supported and managed
by the INET administrative authority. Cost for purchasing,
configuring and managing Servers is nominal especially for virtual
Servers hosted in the cloud.
AERO Proxys are most often standard dedicated server platforms with
one network interface connected to the ANET and a second interface
Templin Expires May 6, 2021 [Page 78]
Internet-Draft AERO November 2020
connected to an INET. As with Servers, the only requirements are
that they can run the AERO user-level code and have at least one
interface connection to the INET. Proxys must be provisioned,
supported and managed by the ANET administrative authority. Cost for
purchasing, configuring and managing Proxys is nominal, and borne by
the ANET administrative authority.
AERO Relays can be any dedicated server or COTS router platform
connected to INETs and/or EUNs. The Relay connects to the OMNI link
and engages in eBGP peering with one or more Bridges as a stub AS.
The Relay then injects its MNPs and/or non-MNP prefixes into the BGP
routing system, and provisions the prefixes to its downstream-
attached networks. The Relay can perform ROS/ROR services the same
as for any Server, and can route between the MNP and non-MNP address
spaces.
A.5. AERO Server Failure Implications
AERO Servers may appear as a single point of failure in the
architecture, but such is not the case since all Servers on the link
provide identical services and loss of a Server does not imply
immediate and/or comprehensive communication failures. Although
Clients typically associate with a single Server at a time, Server
failure is quickly detected and conveyed by Bidirectional Forward
Detection (BFD) and/or proactive NUD allowing Clients to migrate to
new Servers.
If a Server fails, ongoing packet forwarding to Clients will continue
by virtue of the asymmetric neighbor cache entries that have already
been established in route optimization sources (ROSs). If a Client
also experiences mobility events at roughly the same time the Server
fails, unsolicited NA messages may be lost but proxy neighbor cache
entries in the DEPARTED state will ensure that packet forwarding to
the Client's new locations will continue for up to DepartTime
seconds.
If a Client is left without a Server for an extended timeframe (e.g.,
greater than ReachableTime seconds) then existing asymmetric neighbor
cache entries will eventually expire and both ongoing and new
communications will fail. The original source will continue to
retransmit until the Client has established a new Server
relationship, after which time continuous communications will resume.
Therefore, providing many Servers on the link with high availability
profiles provides resilience against loss of individual Servers and
assurance that Clients can establish new Server relationships quickly
in event of a Server failure.
Templin Expires May 6, 2021 [Page 79]
Internet-Draft AERO November 2020
A.6. AERO Client / Server Architecture
The AERO architectural model is client / server in the control plane,
with route optimization in the data plane. The same as for common
Internet services, the AERO Client discovers the addresses of AERO
Servers and selects one Server to connect to. The AERO service is
analogous to common Internet services such as google.com, yahoo.com,
cnn.com, etc. However, there is only one AERO service for the link
and all Servers provide identical services.
Common Internet services provide differing strategies for advertising
server addresses to clients. The strategy is conveyed through the
DNS resource records returned in response to name resolution queries.
As of January 2020 Internet-based 'nslookup' services were used to
determine the following:
o When a client resolves the domainname "google.com", the DNS always
returns one A record (i.e., an IPv4 address) and one AAAA record
(i.e., an IPv6 address). The client receives the same addresses
each time it resolves the domainname via the same DNS resolver,
but may receive different addresses when it resolves the
domainname via different DNS resolvers. But, in each case,
exactly one A and one AAAA record are returned.
o When a client resolves the domainname "ietf.org", the DNS always
returns one A record and one AAAA record with the same addresses
regardless of which DNS resolver is used.
o When a client resolves the domainname "yahoo.com", the DNS always
returns a list of 4 A records and 4 AAAA records. Each time the
client resolves the domainname via the same DNS resolver, the same
list of addresses are returned but in randomized order (i.e.,
consistent with a DNS round-robin strategy). But, interestingly,
the same addresses are returned (albeit in randomized order) when
the domainname is resolved via different DNS resolvers.
o When a client resolves the domainname "amazon.com", the DNS always
returns a list of 3 A records and no AAAA records. As with
"yahoo.com", the same three A records are returned from any
worldwide Internet connection point in randomized order.
The above example strategies show differing approaches to Internet
resilience and service distribution offered by major Internet
services. The Google approach exposes only a single IPv4 and a
single IPv6 address to clients. Clients can then select whichever IP
protocol version offers the best response, but will always use the
same IP address according to the current Internet connection point.
This means that the IP address offered by the network must lead to a
Templin Expires May 6, 2021 [Page 80]
Internet-Draft AERO November 2020
highly-available server and/or service distribution point. In other
words, resilience is predicated on high availability within the
network and with no client-initiated failovers expected (i.e., it is
all-or-nothing from the client's perspective). However, Google does
provide for worldwide distributed service distribution by virtue of
the fact that each Internet connection point responds with a
different IPv6 and IPv4 address. The IETF approach is like google
(all-or-nothing from the client's perspective), but provides only a
single IPv4 or IPv6 address on a worldwide basis. This means that
the addresses must be made highly-available at the network level with
no client failover possibility, and if there is any worldwide service
distribution it would need to be conducted by a network element that
is reached via the IP address acting as a service distribution point.
In contrast to the Google and IETF philosophies, Yahoo and Amazon
both provide clients with a (short) list of IP addresses with Yahoo
providing both IP protocol versions and Amazon as IPv4-only. The
order of the list is randomized with each name service query
response, with the effect of round-robin load balancing for service
distribution. With a short list of addresses, there is still
expectation that the network will implement high availability for
each address but in case any single address fails the client can
switch over to using a different address. The balance then becomes
one of function in the network vs function in the end system.
The same implications observed for common highly-available services
in the Internet apply also to the AERO client/server architecture.
When an AERO Client connects to one or more ANETs, it discovers one
or more AERO Server addresses through the mechanisms discussed in
earlier sections. Each Server address presumably leads to a fault-
tolerant clustering arrangement such as supported by Linux-HA,
Extended Virtual Synchrony or Paxos. Such an arrangement has
precedence in common Internet service deployments in lightweight
virtual machines without requiring expensive hardware deployment.
Similarly, common Internet service deployments set service IP
addresses on service distribution points that may relay requests to
many different servers.
For AERO, the expectation is that a combination of the Google/IETF
and Yahoo/Amazon philosophies would be employed. The AERO Client
connects to different ANET access points and can receive 1-2 Server
LLAs at each point. It then selects one AERO Server address, and
engages in RS/RA exchanges with the same Server from all ANET
connections. The Client remains with this Server unless or until the
Server fails, in which case it can switch over to an alternate
Server. The Client can likewise switch over to a different Server at
any time if there is some reason for it to do so. So, the AERO
Templin Expires May 6, 2021 [Page 81]
Internet-Draft AERO November 2020
expectation is for a balance of function in the network and end
system, with fault tolerance and resilience at both levels.
Appendix B. Change Log
<< RFC Editor - remove prior to publication >>
Changes from draft-templin-intarea-6706bis-61 to draft-templin-
intrea-6706bis-62:
o New sub-section on OMNI Neighbor Interface Attributes
Changes from draft-templin-intarea-6706bis-59 to draft-templin-
intrea-6706bis-60:
o Removed all references to S/TLLAO - all Interface Attributes are
now maintained completely in the OMNI option.
Changes from draft-templin-intarea-6706bis-58 to draft-templin-
intrea-6706bis-59:
o The term "Relay" used in older draft versions is now "Bridge".
"Relay" now refers to what was formally called: "Gateway".
o Fine-grained cleanup of Forwarding Algorithm; IPv6 ND message
addressing; OMNI Prefix Lengths, etc.
Changes from draft-templin-intarea-6706bis-54 to draft-templin-
intrea-6706bis-55:
o Updates on Segment Routing and S/TLLAO contents.
o Various editorials and addressing cleanups.
Changes from draft-templin-intarea-6706bis-52 to draft-templin-
intrea-6706bis-53:
o Normative reference to the OMNI spec, and remove portions that are
already specified in OMNI.
o Renamed "AERO interface/link" to "OMIN interface/link" throughout
the document.
o Truncated obsolete back section matter.
Templin Expires May 6, 2021 [Page 82]
Internet-Draft AERO November 2020
Author's Address
Fred L. Templin (editor)
Boeing Research & Technology
P.O. Box 3707
Seattle, WA 98124
USA
Email: fltemplin@acm.org
Templin Expires May 6, 2021 [Page 83]