]> Thoughtspot

3rd Floor, Indiqube Orion, 24th Main Rd, Garden Layout, HSR Layout Bangalore, KARNATAKA 560 102 India shwetha.bhandari@thoughtspot.com

Cisco Systems, Inc.

Hansaallee 249, 3rd Floor DUESSELDORF 40549 Germany fbrockne@cisco.com

Transport Area ippm In-situ Operations, Administration, and Maintenance (IOAM) records operational and telemetry information in the packet while the packet traverses a path between two points in the network. This document outlines how IOAM data fields are encapsulated in IPv6.

In-situ Operations, Administration, and Maintenance (IOAM) records operational and telemetry information in the packet while the packet traverses a path between two points in the network. IOAM concepts and associated nomenclature, as well as IOAM data fields are defined in . This document outlines how IOAM data fields are encapsulated in IPv6 and discusses deployment requirements for networks that use IPv6-encapsulated IOAM data fields. The terms "encapsulation" and "decapsulation" are used in this document in the same way as in : An IOAM encapsulating node incorporates one or more IOAM-Option-Types into packets. An IOAM decapsulating node removes IOAM-Option-Type(s) from packets.

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 when, and only when, they appear in all capitals, as shown here.

Abbreviations used in this document: Edge-to-Edge In-situ Operations, Administration, and Maintenance as defined in Operations, Administration, and Maintenance Proof of Transit

IOAM in IPv6 is used to enhance diagnostics of IPv6 networks. It complements other mechanisms designed to enhance diagnostics of IPv6 networks, such as the IPv6 Performance and Diagnostic Metrics Destination Option described in . At the time this document was written, several implementations of IOAM for IPv6 exist, e.g., IOAM for IPv6 in the Linux Kernel (supported from Kernel version 5.15 onwards IPv6 IOAM in Linux Kernel), IOAM for IPv6 in VPP . IOAM data fields can be encapsulated with two types of extension headers in IPv6 packets - either the hop-by-hop options header or the destination options header. Multiple options with the same option type MAY appear in the same hop-by-hop options or destination options header, with distinct content. An IPv6 packet carrying IOAM data in an extension header can have other extension headers, compliant with . IPv6 hop-by-hop and destination option format for carrying IOAM data fields:

8-bit option type identifier as defined in . 8-bit unsigned integer. Length of this option, in octets, not including the first 2 octets. 8-bit field MUST be set to zero by the source. Abbreviated to "IOAM-Opt-Type" in the diagram above: 8-bit field as defined in section 4.1 of . Variable-length field. Option-Type-specific data. IOAM Option data is inserted as follows: Pre-allocated Trace Option: The IOAM Preallocated Trace Option-Type defined in Section 4.4 of is represented as an IPv6 option in the hop-by-hop extension header: TBD_1_1 8-bit identifier of the IPv6 Option Type for IOAM. IOAM Pre-allocated Trace Option-Type. Proof of Transit Option: The IOAM POT Option-Type defined in Section 4.5 of is represented as an IPv6 option in the hop-by-hop extension header: TBD_1_1 8-bit identifier of the IPv6 Option Type for IOAM. IOAM POT Option-Type. Edge to Edge Option: The IOAM E2E option defined in Section 4.6 is represented as an IPv6 option in destination extension header: TBD_1_0 8-bit identifier of the IPv6 Option Type for IOAM. IOAM E2E Option-Type. Direct Export (DEX) Option: The IOAM Direct Export Option-Type defined in Section 3.2 of is represented as an IPv6 option in the hop-by-hop extension header: TBD_1_0 8-bit identifier of the IPv6 Option Type for IOAM. IOAM Direct Export (DEX) Option-Type. All the IOAM IPv6 options defined here have alignment requirements. Specifically, they all require 4n alignment. This ensures that fields specified in are aligned at a multiple-of-4 offset from the start of the hop-by-hop and destination options header. IPv6 options can have a maximum length of 255 octets. Consequently, the total length of IOAM Option-Types including all data fields is also limited to 255 octets when encapsulated into IPv6.

IOAM deployments in IPv6 networks MUST take the following considerations and requirements into account: IOAM MUST be deployed in an IOAM-Domain. An IOAM-Domain is a set of nodes that use IOAM. An IOAM-Domain is bounded by its perimeter or edge. The set of nodes forming an IOAM-Domain may be connected to the same physical infrastructure (e.g., a service provider's network). They may also be remotely connected to each other (e.g., an enterprise VPN or an overlay). It is expected that all nodes in an IOAM-Domain are managed by the same administrative entity. Please refer to ) for more details on IOAM-Domains. Implementations of IOAM MUST ensure that the addition of IOAM data fields does not alter the way routers forward packets or the forwarding decisions they make. Packets with added IOAM information must follow the same path within the domain as an identical packet without IOAM information would, even in the presence of Equal-Cost Multi-Path (ECMP). This behavior is important for deployments where IOAM data fields are only added "on-demand". Implementations of IOAM MUST ensure that ECMP behavior for packets with and without IOAM data fields is the same. In order for IOAM to work in IPv6 networks, IOAM MUST be explicitly enabled per interface on every node within the IOAM domain. Unless a particular interface is explicitly enabled (i.e., explicitly configured) for IOAM, a router MUST ignore IOAM Options. In order to maintain the integrity of packets in an IOAM domain, the Maximum Transmission Unit (MTU) of transit routers and switches must be configured to a value that does not lead to an ICMP Packet Too Big error message being sent to the originator and the packet being dropped. The PMTU tolerance range must be identified and IOAM encapsulation operations or data field insertion must not exceed this range. Control of the MTU is critical to the proper operation of IOAM. The PMTU tolerance must be identified through configuration and IOAM operations must not exceed the packet size beyond PMTU. precludes insertion of IOAM data directly into the original IPv6 header of in-flight packets. IOAM deployments which do not encapsulate/decapsulate IOAM on the host but desire to encapsulate/decapsulate IOAM on transit nodes MUST add an additional IPv6 header to the original packet. IOAM data is added to this additional IPv6 header.

For deployments where the IOAM domain is bounded by hosts, hosts will perform the operation of IOAM data field encapsulation and decapsulation, i.e., hosts will place the IOAM data fields directly in the IPv6 header or remove the IOAM data fields directly from the IPv6 header. IOAM data is carried in IPv6 packets as hop-by-hop or destination options as specified in this document.

For deployments where the IOAM domain is bounded by network devices, network devices such as routers form the edge of an IOAM domain. Network devices will perform the operation of IOAM data field encapsulation and decapsulation. Network devices will encapsulate IOAM data fields in an additional, outer, IPv6 header which carries the IOAM data fields.

This document describes the encapsulation of IOAM data fields in IPv6. For general IOAM security considerations, see . Security considerations of the specific IOAM data fields for each case (i.e., Trace, Proof of Transit, and E2E) are also described and defined in . As this document describes new options for IPv6, the security considerations of and apply. From a network-protection perspective, there is an assumed trust model such that any node that adds IOAM to a packet, removes IOAM from a packet, or modifies IOAM data fields of a packet is assumed to be allowed to do so. By default, packets that include IPv6 extension headers with IOAM information MUST NOT be leaked through the boundaries of the IOAM-Domain. IOAM-Domain boundary routers MUST filter any incoming traffic from outside the IOAM-Domain that contains IPv6 extension headers with IOAM information. IOAM-Domain boundary routers MUST also filter any outgoing traffic leaving the IOAM-Domain that contains IPv6 extension headers with IOAM information. In the general case, an IOAM node only adds, removes, or modifies an IPv6 extension header with IOAM information, if the directive to do so comes from a trusted source and the directive is validated. Problems may occur if the above behaviors are not implemented or if the assumed trust model is violated (e.g., through a security breach). In addition to the security considerations discussed in , the security considerations associated with IPv6 extension headers listed in apply.

The network devices in an IOAM-Domain are trusted to add, update and remove IOAM options according to the constraints specified in . IOAM domain does not rely on the Authentication Header (AH) as defined in to secure IOAM options. The use of IOAM options with AH and its processing is not defined in this document. Future documents may define the use of IOAM with AH and its processing.

This draft requests the following IPv6 Option Type assignments from the destination options and hop-by-hop options sub-registry of Internet Protocol Version 6 (IPv6) Parameters. http://www.iana.org/assignments/ipv6-parameters/ipv6- parameters.xhtml#ipv6-parameters-2

The authors would like to thank Tom Herbert, Eric Vyncke, Nalini Elkins, Srihari Raghavan, Ranganathan T S, Karthik Babu Harichandra Babu, Akshaya Nadahalli, Stefano Previdi, Hemant Singh, Erik Nordmark, LJ Wobker, Mark Smith, Andrew Yourtchenko and Justin Iurman for the comments and advice. For the IPv6 encapsulation, this document leverages concepts described in . The authors would like to acknowledge the work done by the author Hiroshi Kitamura and people involved in writing it.

&RFC2119; &RFC8174; &RFC9197; &RFC9326; &RFC8200; &RFC8250; &RFC4302; &RFC9098;

This document was the collective effort of several authors. The text and content were contributed by the editors and the co-authors listed below. The contact information of the co-authors appears at the end of this document. Carlos Pignataro Hannes Gredler John Leddy Stephen Youell Tal Mizrahi Aviv Kfir Barak Gafni Petr Lapukhov Mickey Spiegel Suresh Krishnan Rajiv Asati Mark Smith