Internet-Draft | Ping Enabled IOAM Capabilities | November 2022 |
Min, et al. | Expires 10 May 2023 | [Page] |
This document describes a generic format for use in echo request/reply mechanisms, which can be used within an In situ Operations, Administration, and Maintenance (IOAM) domain, allowing the IOAM encapsulating node to discover the enabled IOAM capabilities of each IOAM transit and IOAM decapsulating node. The generic format is intended to be used with a variety of data planes such as IPv6, MPLS, Service Function Chain (SFC) and Bit Index Explicit Replication (BIER).¶
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 10 May 2023.¶
Copyright (c) 2022 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 Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶
In situ Operations, Administration, and Maintenance (IOAM) ([RFC9197] [I-D.ietf-ippm-ioam-direct-export]) defines data fields that record OAM information within the packet while the packet traverses a particular network domain, called an IOAM domain. IOAM can complement or replace other OAM mechanisms, such as ICMP or other types of probe packets.¶
As specified in [RFC9197], within the IOAM domain, the IOAM data may be updated by network nodes that the packet traverses. The device which adds an IOAM header to the packet is called an "IOAM encapsulating node". In contrast, the device which removes an IOAM header is referred to as an "IOAM decapsulating node". Nodes within the domain that are aware of IOAM data and read and/or write and/or process IOAM data are called "IOAM transit nodes". IOAM encapsulating or decapsulating nodes can also serve as IOAM transit nodes at the same time. IOAM encapsulating or decapsulating nodes are also referred to as IOAM domain edge devices, which can be hosts or network devices. [RFC9197] defines four IOAM option types, and [I-D.ietf-ippm-ioam-direct-export] introduces a new IOAM option type called the Direct Export (DEX) Option-Type, which is different from the other four IOAM option types defined in [RFC9197] on how to collect the operational and telemetry information defined in [RFC9197].¶
As specified in [RFC9197], IOAM is focused on "limited domains" as defined in [RFC8799]. In a limited domain, a control entity that has control over every IOAM device may be deployed. If that's the case, the control entity can provision both the explicit transport path and the IOAM header applied to data packet at every IOAM encapsulating node.¶
In a case when a control entity that has control over every IOAM device is not deployed in the IOAM domain, the IOAM encapsulating node needs to discover the enabled IOAM capabilities at the IOAM transit and decapsulating nodes. For example, what types of IOAM tracing data can be added or exported by the transit nodes along the transport path of the data packet IOAM is applied to. The IOAM encapsulating node can then add the correct IOAM header to the data packet according to the discovered IOAM capabilities. Specifically, the IOAM encapsulating node first identifies the types and lengths of IOAM options included in the IOAM data fields according to the discovered IOAM capabilities. Then the IOAM encapsulating node can add the IOAM header to the data packet based on the identified types and lengths of IOAM options included in the IOAM data fields. The IOAM encapsulating node may use NETCONF/YANG or IGP to discover these IOAM capabilities. However, NETCONF/YANG or IGP has some limitations:¶
This document specifies formats and objects that can be used in the extension of echo request/reply mechanisms used in IPv6 (including Segment Routing with IPv6 data plane (SRv6)), MPLS (including Segment Routing with MPLS data plane (SR-MPLS)), SFC and BIER environments, which can be used within the IOAM domain, allowing the IOAM encapsulating node to discover the enabled IOAM capabilities of each IOAM transit and IOAM decapsulating node.¶
The following documents contain references to the echo request/reply mechanisms used in IPv6 (including SRv6), MPLS (including SR-MPLS), SFC and BIER environments:¶
It is expected that the specification of the instantiation of each of these extensions will be done in the form of an RFC jointly designed by the working group that develops or maintains the echo request/reply protocol and the IETF IP Performance Measurement (IPPM) Working Group.¶
Note that in this document the echo request/reply mechanism used in IPv6 does not mean ICMPv6 Echo Request/Reply [RFC4443], but means IPv6 Node Information Query/Reply [RFC4620].¶
Fate sharing is a common requirement for all kinds of active OAM packets, echo request is among them, in this document that means echo request is required to traverse a path of IOAM data packet. This requirement can be achieved by, e.g., applying same explicit path or ECMP processing to both echo request and IOAM data packet. Specific to apply same ECMP processing to both echo request and IOAM data packet, one possible way is to populate the same value(s) of ECMP affecting field(s) in the echo request.¶
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.¶
BIER: Bit Index Explicit Replication¶
BGP: Border Gateway Protocol¶
DEX: Direct Export¶
ECMP: Equal-Cost Multipath¶
E2E: Edge to Edge¶
ICMP: Internet Control Message Protocol¶
IGP: Interior Gateway Protocol¶
IOAM: In situ Operations, Administration, and Maintenance¶
LSP: Label Switched Path¶
MPLS: Multi-Protocol Label Switching¶
MTU: Maximum Transmission Unit¶
NTP: Network Time Protocol¶
OAM: Operations, Administration, and Maintenance¶
PCEP: Path Computation Element (PCE) Communication Protocol¶
POSIX: Portable Operating System Interface¶
POT: Proof of Transit¶
PTP: Precision Time Protocol¶
SR-MPLS: Segment Routing with MPLS data plane¶
SRv6: Segment Routing with IPv6 data plane¶
SFC: Service Function Chain¶
TTL: Time to Live, this is also the Hop Limit field in the IPv6 header¶
For echo request, IOAM Capabilities Query uses a container which has the following format:¶
When this container is present in the echo request sent by an IOAM encapsulating node, that means the IOAM encapsulating node requests the receiving node to reply with its enabled IOAM capabilities. If there is no IOAM capability to be reported by the receiving node, then this container MUST be ignored by the receiving node, which means the receiving node MUST send an echo reply without IOAM capabilities or no echo reply, in the light of whether the echo request includes other containers than the IOAM Capabilities Query Container. A list of IOAM Namespace-IDs (one or more Namespace-IDs) MUST be included in this container in the echo request, and if present, the Default-Namespace-ID 0x0000 MUST be placed at the beginning of the list of IOAM Namespace-IDs. The IOAM encapsulating node requests only the enabled IOAM capabilities that match one of the Namespace-IDs. Inclusion of the Default-Namespace-ID 0x0000 elicits replies only for capabilities that are configured with the Default-Namespace-ID 0x0000.The Namespace-ID has the same definition as what's specified in Section 4.3 of [RFC9197].¶
The IOAM Capabilities Query Container has a container header that is used to identify the type and optionally length of the container payload, and the container payload (List of IOAM Namespace-IDs) is zero-padded to align to a 4-octet boundary. Note that since the Default-Namespace-ID of 0x0000 is mandated to appear first in the list, if it appears any trailing 0x0000 octets must therefore be padding and MUST be disregarded.¶
The length, structure, and definition of the IOAM Capabilities Query Container Header depends on the specific deployment environment.¶
For echo reply, IOAM Capabilities Response uses a container which has the following format:¶
When this container is present in the echo reply sent by an IOAM transit node or IOAM decapsulating node, that means the IOAM function is enabled at this node, and this container contains the enabled IOAM capabilities of the sender. A list of IOAM capabilities objects (one or more objects) which contains the enabled IOAM capabilities MUST be included in this container of echo reply except the sender encounters an error (e.g., no matched Namespace-ID).¶
The IOAM Capabilities Response Container has a container header that is used to identify the type and optionally length of the container payload. The container header MUST be defined such that it falls on a four-octet boundary.¶
The length, structure, and definition of the IOAM Capabilities Response Container Header depends on the specific deployment environment.¶
Based on the IOAM data fields defined in [RFC9197] and [I-D.ietf-ippm-ioam-direct-export], six types of objects are defined in this document. The same type of object MAY be present in the IOAM Capabilities Response Container more than once, only if with a different Namespace-ID.¶
Similar to the container, each object has an object header that is used to identify the type and length of the object payload. The object payload MUST be defined such that it falls on a four-octet boundary.¶
The length, structure, and definition of Object Header depends on the specific deployment environment.¶
When this Object is present in the IOAM Capabilities Response Container, that means the sending node is an IOAM transit node and the IOAM pre-allocated tracing function is enabled at this IOAM transit node.¶
IOAM-Trace-Type field has the same definition as what's specified in Section 4.4 of [RFC9197].¶
Reserved field is reserved for future use and MUST be set to zero, and MUST be ignored when non-zero.¶
W flag indicates whether Ingress_if_id is in short or wide format. The W-bit is set if the Ingress_if_id is in wide format. The W-bit is clear if the Ingress_if_id is in short format.¶
Namespace-ID field has the same definition as what's specified in Section 4.3 of [RFC9197], it MUST be one of the Namespace-IDs listed in the IOAM Capabilities Query Object of the echo request.¶
Ingress_MTU field has 16 bits and specifies the MTU (in octets) of the ingress interface from which the sending node received echo request.¶
Ingress_if_id field has 16 bits (in short format) or 32 bits (in wide format) and specifies the identifier of the ingress interface from which the sending node received echo request. If the W-bit is cleared that indicates Ingress_if_id field has 16 bits, then the 16 bits following the Ingress_if_id field are reserved for future use and MUST be set to zero, and MUST be ignored when non-zero.¶
When this Object is present in the IOAM Capabilities Response Container, that means the sending node is an IOAM transit node and the IOAM incremental tracing function is enabled at this IOAM transit node.¶
IOAM-Trace-Type field has the same definition as what's specified in Section 4.4 of [RFC9197].¶
Reserved field is reserved for future use and MUST be set to zero, and MUST be ignored when non-zero.¶
W flag indicates whether Ingress_if_id is in short or wide format. The W-bit is set if the Ingress_if_id is in wide format. The W-bit is clear if the Ingress_if_id is in short format.¶
Namespace-ID field has the same definition as what's specified in Section 4.3 of [RFC9197], it MUST be one of the Namespace-IDs listed in the IOAM Capabilities Query Object of the echo request.¶
Ingress_MTU field has 16 bits and specifies the MTU (in octets) of the ingress interface from which the sending node received echo request.¶
Ingress_if_id field has 16 bits (in short format) or 32 bits (in wide format) and specifies the identifier of the ingress interface from which the sending node received echo request. If the W-bit is cleared that indicates Ingress_if_id field has 16 bits, then the 16 bits following the Ingress_if_id field are reserved for future use and MUST be set to zero, and MUST be ignored when non-zero.¶
When this Object is present in the IOAM Capabilities Response Container, that means the sending node is an IOAM transit node and the IOAM Proof of Transit function is enabled at this IOAM transit node.¶
Namespace-ID field has the same definition as what's specified in Section 4.3 of [RFC9197], it MUST be one of the Namespace-IDs listed in the IOAM Capabilities Query Object of the echo request.¶
IOAM-POT-Type field has the same definition as what's specified in Section 4.5 of [RFC9197].¶
SoP (Size of POT) field has two bits, which means the size of "PktID" and "Cumulative" data that are specified in Section 4.5 of [RFC9197]. This document defines SoP as follows:¶
Reserved field is reserved for future use and MUST be set to zero, and MUST be ignored when non-zero.¶
When this Object is present in the IOAM Capabilities Response Container, that means the sending node is an IOAM decapsulating node and IOAM edge-to-edge function is enabled at this IOAM decapsulating node.¶
Namespace-ID field has the same definition as what's specified in Section 4.3 of [RFC9197], it MUST be one of the Namespace-IDs listed in the IOAM Capabilities Query Object of the echo request.¶
IOAM-E2E-Type field has the same definition as what's specified in Section 4.6 of [RFC9197].¶
TSF field specifies the timestamp format used by the sending node. Aligned with three possible timestamp formats specified in Section 5 of [RFC9197], this document defines TSF as follows:¶
Reserved field is reserved for future use and MUST be set to zero, and MUST be ignored when non-zero.¶
When this Object is present in the IOAM Capabilities Response Container, that means the sending node is an IOAM transit node and the IOAM direct exporting function is enabled at this IOAM transit node.¶
IOAM-Trace-Type field has the same definition as what's specified in Section 3.2 of [I-D.ietf-ippm-ioam-direct-export].¶
Namespace-ID field has the same definition as what's specified in Section 4.3 of [RFC9197], it MUST be one of the Namespace-IDs listed in the IOAM Capabilities Query Object of the echo request.¶
Reserved field is reserved for future use and MUST be set to zero, and MUST be ignored when non-zero.¶
When this Object is present in the IOAM Capabilities Response Container, that means the sending node is an IOAM decapsulating node. Unless the IOAM Edge-to-Edge Capabilities Object is present, which also indicates that the sending node is an IOAM decapsulating node, the End-of-Domain Object MUST be present in the IOAM Capabilities Response Container sent by an IOAM decapsulating node. When the IOAM edge-to-edge function is enabled at the IOAM decapsulating node, it's RECOMMENDED to include only the IOAM Edge-to-Edge Capabilities Object but not the IOAM End-of-Domain Object.¶
Namespace-ID field has the same definition as what's specified in Section 4.3 of [RFC9197], it MUST be one of the Namespace-IDs listed in the IOAM Capabilities Query Container.¶
Once the IOAM encapsulating node is triggered to discover the enabled IOAM capabilities of each IOAM transit and IOAM decapsulating node, the IOAM encapsulating node will send echo requests that include the IOAM Capabilities Query Container. First, with TTL equal to 1 to reach the closest node, which may be an IOAM transit node or not. Then with TTL equal to 2 to reach the second-nearest node, which also may be an IOAM transit node or not. And further, increasing by 1 the TTL every time the IOAM encapsulating node sends a new echo request, until the IOAM encapsulating node receives an echo reply sent by the IOAM decapsulating node, which contains the IOAM Capabilities Response Container including the IOAM Edge-to-Edge Capabilities Object or the IOAM End-of-Domain Object. As a result, the echo requests sent by the IOAM encapsulating node will reach all nodes one by one along the transport path of IOAM data packet. Alternatively, if the IOAM encapsulating node knows precisely all the IOAM transit and IOAM decapsulating nodes beforehand, once the IOAM encapsulating node is triggered to discover the enabled IOAM capabilities, it can send an echo request to each IOAM transit and IOAM decapsulating node directly, without TTL expiration.¶
The IOAM encapsulating node may be triggered by the device administrator, the network management system, the network controller, or data traffic. The specific triggering mechanisms are outside the scope of this document.¶
Each IOAM transit and IOAM decapsulating node that receives an echo request containing the IOAM Capabilities Query Container will send an echo reply to the IOAM encapsulating node. For the echo reply, there is an IOAM Capabilities Response Container containing one or more Objects. The IOAM Capabilities Query Container of the echo request would be ignored by the receiving node unaware of IOAM.¶
Note that the mechanism defined in this document applies to all kinds of IOAM option types, whether the four types of IOAM option defined in [RFC9197] or the DEX type of IOAM option defined in [I-D.ietf-ippm-ioam-direct-export], specifically, when applied to the IOAM DEX option, it allows the IOAM encapsulating node to discover which nodes along the transport path support IOAM direct exporting and which trace data types are supported to be directly exported at these nodes.¶
This document requests the following IANA Actions.¶
IANA is requested to create a registry group named "In-Situ OAM (IOAM) Capabilities Parameters".¶
This group will include the following registries:¶
New registries in this group can be created via RFC Required process as per [RFC8126].¶
The subsequent subsections detail the registries herein contained.¶
Considering the Containers/Objects defined in this document would be carried in different types of Echo Request/Reply messages, such as ICMPv6 or LSP Ping, it is intended that the registries for Container/Object Type would be requested in subsequent documents.¶
This registry defines 4 code points for the IOAM SoP Capability field for identifying the size of "PktID" and "Cumulative" data as explained in Section 4.5 of [RFC9197]. The following code points are defined in this document:¶
SoP Description ---- ----------- 0b00 64-bit "PktID" and 64-bit "Cumulative" data¶
0b01 - 0b11 are available for assignment via IETF Review process as per [RFC8126].¶
This registry defines 4 code points for the IOAM TSF Capability field of identifying the timestamp format as explained in Section 5 of [RFC9197]. The following code points are defined in this document:¶
TSF Description ---- ----------- 0b00 PTP Truncated Timestamp Format 0b01 NTP 64-bit Timestamp Format 0b10 POSIX-based Timestamp Format 0b11 Reserved for future standardization¶
0b11 is available for assignment via IETF Review process as per [RFC8126].¶
Overall, the security needs for IOAM capabilities query mechanisms used in different environments are similar.¶
To avoid potential Denial-of-Service (DoS) attacks, it is RECOMMENDED that implementations apply rate-limiting to incoming echo requests and replies.¶
To protect against unauthorized sources using echo request messages to obtain IOAM Capabilities information, implementations MUST provide a means of checking the source addresses of echo request messages against an access list before accepting the message.¶
A deployment MUST ensure that border filtering drops inbound echo requests with an IOAM Capabilities Container Header from outside of the domain, and drops outbound echo request/replies with IOAM Capabilities Headers leaving the domain.¶
A deployment MUST support the configuration option to enable/disable the IOAM Capabilities Discovery feature defined in this document. By default, the IOAM Capabilities Discovery feature MUST be disabled.¶
The integrity protection on IOAM Capabilities information carried in echo reply messages can be achieved by the underlying transport. For example, if the environment is an IPv6 network, the IP Authentication Header [RFC4302] or IP Encapsulating Security Payload Header [RFC4303] can be used.¶
The collected IOAM Capabilities information by queries may be considered confidential. An implementation can use secure underlying transport of echo request/reply to provide privacy protection. For example, if the environment is an IPv6 network, confidentiality can be achieved by using the IP Encapsulating Security Payload Header [RFC4303].¶
An implementation can also directly secure the data carried in echo requests and replies if needed, the specific mechanism on how to secure the data is beyond the scope of this document.¶
An implementation can also check whether the fields in received echo requests and replies strictly conform to the specifications, e.g., whether the list of IOAM Namespace-IDs includes duplicate entries, whether the received Namespace-ID is an operator-assigned or IANA-assigned one, once a check fails, an exception event indicating the checked field should be reported to the management.¶
Except for what's described above, the security issues discussed in [RFC9197] provide a good guidance on implementation of this specification.¶
The authors would like to acknowledge Tianran Zhou, Dhruv Dhody, Frank Brockners, Cheng Li, Gyan Mishra, Marcus Ihlar, Martin Duke, Chris Lonvick, Eric Vyncke, Alvaro Retana, Paul Wouters, Roman Danyliw, Lars Eggert, Warren Kumari, John Scudder, Robert Wilton, Erik Kline and Zaheduzzaman Sarker for their careful review and helpful comments.¶
The authors appreciate the f2f discussion with Frank Brockners on this document.¶
The authors would like to acknowledge Tommy Pauly and Ian Swett for their good suggestion and guidance.¶