]> ZTE Corporation

fu.huakai@zte.com.cn

ZTE Corporation

xiong.quan@zte.com.cn

China Mobile

liubo@chinamobile.com

China Mobile

lizhenqiang@chinamobile.com

CATS This document describes the Operations, Administration, and Maintenance (OAM) framework and requirements for Computing-Aware Traffic Steering (CATS). The framework defines the CATS OAM layering model and functional components. It also specifies the requirements to enable fault management and performance monitoring for CATS end-to-end connections encompassing clients, network paths, and service instances.

Introduction As described in , edge computing provides lower response time and higher transmission rate than cloud computing by moving computing resources to the network edge. To meet the requirements of users that are highly distributive, service providers deploy the same type of service instances at multiple edge sites, which involves steering traffic from clients to the most appropriate computing instance. Computing-Aware Traffic Steering (CATS) provides a traffic engineering approach that incorporates the dynamic states of both computing and network resources to optimize service instance selection. While such policies rely on multi-dimensional metrics to achieve service assurance, existing network-centric OAM technologies are insufficient as they focus solely on infrastructure-layer maintenance and fail to provide end-to-end visibility from the client to the service instance. Consequently, a dedicated CATS OAM framework is required to bridge the gap between network reachability and service availability, transforming CATS from a theoretical steering logic into a manageable, carrier-grade service capable of sustaining performance in distributed computing environments. To this end, and aligned with the architecture defined in , this document specifies the OAM framework and functional requirements for CATS. It establishes a layered OAM model and defines the necessary functional components to monitor the system. Furthermore, it outlines the requirements for fault management and performance monitoring across the entire CATS service chain, covering the connection from the client through the network to the final service instance.

Requirements Language 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.

Terminology This document makes use of the terms defined in [I-D.ietf-cats- framework].

• FM: Fault Management.
• PM: Performance Monitoring.
• SI-OAM: Service Instance OAM.
• SVC-OAM: Service OAM.

Motivation and Problem Statement Existing Operations, Administration, and Maintenance (OAM) mechanisms, such as those defined in , are primarily optimized for network-layer connectivity verification and path-level performance monitoring. However, the emergence of Computing-Aware Traffic Steering (CATS) introduces a multidimensional decision-making process. This necessitates an OAM framework capable of perceiving both network transport characteristics and service-level operational capabilities. The primary objective of a CATS-specific OAM framework is to ensure that traffic steering decisions are aligned with the real-time operational status of distributed computing resources. This is critical to preventing "service black-holing," a condition where traffic is steered to a service instance that remains reachable at the network layer but is functionally unresponsive or degraded at the application level. In the absence of a dedicated OAM framework for CATS, several critical gaps persist:

• Reachability without Serviceability：Traditional network-layer OAM protocols (e.g., Bidirectional Forwarding Detection (BFD) , Internet Control Message Protocol(ICMP) ) verify the liveness of the routing path and the node's IP interface. They lack the granularity to detect the health of the service instance hosted on that node. Consequently, a CATS Path Selector (C-PS) may continue to steer traffic to a node where the application process has crashed, deadlocked, or exceeded its resource limits, despite the node remaining network-reachable.
• Temporal Staleness of Computing Metrics:Computing metrics (e.g., CPU utilization, active thread counts, and task queue depths) exhibit significantly higher volatility compared to network topology changes. Standard control-plane advertisement or polling mechanisms often lack the necessary frequency to capture these micro-bursts, resulting in "stale" telemetry. This synchronization latency leads to sub-optimal steering decisions, potential traffic flapping, and inconsistent client experiences.
• Lack of Cross-Layer Fault Demarcation：When end-to-end (E2E) service degradation occurs (e.g., increased response latency), current OAM tools cannot natively distinguish between network-induced bottlenecks (e.g., path congestion) and compute-induced bottlenecks (e.g., resource exhaustion at the SI). This ambiguity hampers rapid fault localization, significantly increases the Mean Time to Repair (MTTR), and complicates Service Level Agreement (SLA) enforcement in multi-domain or multi-vendor environments.
• Incommensurability of Heterogeneous Metrics: Since network metrics (e.g., one-way delay, jitter, packet loss) and computing metrics (e.g., instructions per second, memory watermark) characterize distinct dimensions of service delivery, they are often collected via disjoint protocols and lack unified correlation mechanisms, such as high-precision synchronized timestamps. Without a consolidated OAM framework to harmonize these data points, the CATS Path Selector (C-PS) cannot accurately calculate the composite cost, defined as C_{total} = f(C_{network}, C_{computing}), thereby undermining the deterministic nature of path computation and steering efficiency.

CATS OAM Framework

CATS OAM Layering Model The CATS OAM hierarchical model leverages the principles of Maintenance Domain (MD) levels, as defined in IEEE 802.1ag and ITU-T Y.1731, to extend traditional connectivity and performance management into the computing domain. This framework enables CATS-Forwarders and underlying network nodes to perform integrated anomaly detection and performance monitoring. Based on the scope of awareness and functional granularity, the CATS OAM mechanisms are organized into four distinct layers: Link OAM, Path OAM, Instance OAM, and Service OAM. The architecture is illustrated in Figure 1.

• Link OAM: This layer is directed at the detection of physical links or single-hop IP interfaces, representing the foundational underlay infrastructure in CATS framework . It encompasses both the internal links within the operator's infrastructure and the external interfaces interconnecting the network and computing domains. The primary objective is to monitor the operational status between two adjacent devices to ensure fundamental IP reachability. The existing detection tools include IEEE 802.3ah, ICMP , and single-hop BFD .
• Path OAM: This layer focuses on the monitoring of transport paths between an ingress CATS-Forwarder and an egress CATS-Forwarder in CATS framework . As an aggregate transport path typically carries multiple services, Path OAM is critical for ensuring that network-layer faults or resource contention from traffic multiplexing do not degrade specific service performance. Fault detection and performance monitoring are executed at the Label Switched Path (LSP) or Segment Routing (SR) path layer to facilitate rapid service protection. The existing detection mechanisms include ITU-T Y.1711, MPLS Loss and Delay Measurement (LM-DM) , and BFD for LSP.
• Instance OAM: This layer is dedicated to the monitoring of status of computing resource and the operational health of service instances in CATS framework . The metrics collected at this layer, such as CPU load and memory availability, are defined in . Monitoring mechanisms MUST support flexible implementation modes, including a "Push" mode, where computing nodes report dynamic load status in real-time, and a "Pull" mode, where the egress CATS-Forwarder proactively retrieves computing metrics by extending the existing OAM mechanisms.
• Service OAM: This layer provides either end-to-end or segmented visibility into the connectivity and performance between the Ingress CATS-Forwarder and the target Service Instance (SI), as defined in the CATS framework . Beyond basic reachability verification, SVC-OAM is responsible for monitoring service-specific liveness, transaction success rates, and application-layer latency to ensure consistent end-to-end Quality of Experience (QoE). The supported detection mechanisms include, but are not limited to, the Two-Way Active Measurement Protocol (TWAMP) , application-aware probing, and HTTP-based health checks, with extensibility to accommodate emerging application-specific requirements.

CATS OAM Components In accordance with Section 5.2 of the CATS framework, the CATS OAM layering model is designed to flexibly accommodate diverse OAM detection mechanisms. Consequently, this document proposes the following two new components including SI-OAM, and SVC-OAM as Figure 2 shown. |          +--+-----+                         +--------+    |          | SVC_OAM |<------------------------------------->|          +--+-----+                                       | Figure 2: CATS OAM Functional Components ]]>

SI-OAM Component The SI-OAM component is responsible for monitoring the operational status and computing capabilities of individual service instances. It facilitates the granular perception of service instances availability and performance. Its key functions include:

• Status Monitoring: The mechanism for periodically verifying the availability and operational state (e.g., active, inactive, or maintenance mode) of a specific service instance. This process is executed between an Egress CATS-Forwarder and its associated service instance to ensure real-time reachability and prevent traffic from being steered to an unavailable or degraded target.
• Computing Metric Collection: The process of gathering real-time computing resource telemetry from the service instance or its hosting environment. Relevant metrics include, but are not limited to, processing capacity, memory watermark, and internal queuing delay, which characterize the instantaneous capability of the instance.
• Metric Reporting: The capability to provide normalized computing telemetry to the CATS Service Metric Agent (C-SMA). This function ensures that raw resource data is translated into a consistent format to support the dynamic update of computing-aware traffic steering policies within the CATS control plane.

SRV-OAM Component The SVC-OAM component provides end-to-end (E2E) visibility and performance assessment for a CATS service across the entire delivery path, originating from an Ingress CATS-Forwarder to the target service instances. Its key functions include:

• Policy Verification: Ensuring that the actual traffic forwarding path from the Ingress CATS-Router to the selected Service Instance aligns with the steering decisions made by the CATS Path Selector (C-PS).
• Joint Performance Measurement: Measuring E2E performance metrics, such as total latency (the sum of network transport time and service processing time) and jitter, to verify Service Level Agreement (SLA) compliance.
• Multi-Domain Fault Isolation: Correlating Link OAM, Path OAM, Instance OAM, and Service OAM to differentiate whether service degradation is caused by network congestion or computing resource exhaustion.

CATS OAM Requirements This section specifies the OAM requirements for CATS, adhering to the operational and management guidelines defined in . CATS OAM must bridge the gap between traditional network connectivity checks and the awareness of computing resource availability.

Operation The operational layer is responsible for generating and collecting metrics, as well as the real-time reporting of the combined network and computing status.

• O-REQ-1: Multi-Dimensional Status Awareness. The system MUST support the collection of both network metrics (latency, jitter, packet loss) and computing metrics ( e.g., processing capacity, load, and availability) defined in . To balance precision and control plane overhead, the system SHOULD support both periodic and threshold-triggered reporting.
• O-REQ-2: Service ID and Location Mapping. The OAM component MUST maintain the binding between a Service ID (representing the service instance) and its specific network location (Egress CATS-Forwarder). OAM probes MUST be able to target specific service instances to verify that the traffic steering policy correctly reaches the intended destination.
• O-REQ-3: Telemetry Integration with C-SMA. All collected computing metrics SHOULD be reported to the CATS Service Metric Agent (C-SMA). This ensures that the CATS Path Selector (C-PS) has access to synchronized telemetry to perform real-time path computation and selection.

Administration The administrative layer focuses on policy definition, security boundaries, and the configuration of steering behaviors.

• A-REQ-1: Policy-Based Steering Configuration. The system MUST allow administrators to define CATS-specific policies, such as weighting factors for network vs. computing metrics. These policies dictate how the C-PS interprets raw telemetry when calculating the "best" service instance.
• A-REQ-2: Differentiated Monitoring Intensity. Based on the Service Level Agreement (SLA), the system SHOULD support variable OAM intensities. High-priority services (e.g., autonomous driving or remote surgery) may require millisecond-level BFD-like monitoring, while standard web services use lower-frequency heartbeats.
• A-REQ-3: Security and Metric Integrity. Reporting of computing and network metrics MUST be protected via encryption and authentication. This prevents malicious actors from injecting "fake" high-performance metrics to attract and intercept traffic (sinkhole attacks) or triggering oscillations in the CATS-Forwarder.

Maintenance The maintenance layer focuses on fault isolation, performance backtracking, and ensuring the consistency of the steering state.

• M-REQ-1: Joint Fault Demarcation. The system MUST be able to distinguish whether a service failure is caused by a network-layer issue (e.g., connectivity loss between CATS-Forwarders) or computing resource exhaustion at the Service Instance. This requires the correlation of multi-layer OAM data, including Link OAM, Path OAM, Instance OAM, and Service OAM, to accurately isolate the fault domain.
• M-REQ-2: Historical Traceability. The OAM system SHOULD record historical snapshots of both network paths and computing status. This is critical for post-mortem analysis of "flapping" steering decisions where traffic frequently oscillates between different service instances.
• M-REQ-3: Forwarding Plane Consistency Check. The system MUST provide mechanisms to verify that the actual traffic path taken by a packet matches the decision made by the C-PS. Any inconsistency between the intended steering policy and the actual forwarding state MUST trigger an immediate alarm and re-synchronization.

Security Considerations To be discussed in future versions of this document.

IANA Considerations TBD.

Acknowledgements To be added upon contributions, comments and suggestions.

References Normative References In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements. RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings. Segment Routing (SR) leverages the source routing paradigm. A node steers a packet through an ordered list of instructions, called "segments". A segment can represent any instruction, topological or service based. A segment can have a semantic local to an SR node or global within an SR domain. SR provides a mechanism that allows a flow to be restricted to a specific topological path, while maintaining per-flow state only at the ingress node(s) to the SR domain. SR can be directly applied to the MPLS architecture with no change to the forwarding plane. A segment is encoded as an MPLS label. An ordered list of segments is encoded as a stack of labels. The segment to process is on the top of the stack. Upon completion of a segment, the related label is popped from the stack. SR can be applied to the IPv6 architecture, with a new type of routing header. A segment is encoded as an IPv6 address. An ordered list of segments is encoded as an ordered list of IPv6 addresses in the routing header. The active segment is indicated by the Destination Address (DA) of the packet. The next active segment is indicated by a pointer in the new routing header. Segment Routing can be applied to the IPv6 data plane using a new type of Routing Extension Header called the Segment Routing Header (SRH). This document describes the SRH and how it is used by nodes that are Segment Routing (SR) capable. Operations, Administration, and Maintenance (OAM) is a general term that refers to a toolset for fault detection and isolation, and for performance measurement. Over the years, various OAM tools have been defined for various layers in the protocol stack. This document summarizes some of the OAM tools defined in the IETF in the context of IP unicast, MPLS, MPLS Transport Profile (MPLS-TP), pseudowires, and Transparent Interconnection of Lots of Links (TRILL). This document focuses on tools for detecting and isolating failures in networks and for performance monitoring. Control and management aspects of OAM are outside the scope of this document. Network repair functions such as Fast Reroute (FRR) and protection switching, which are often triggered by OAM protocols, are also out of the scope of this document. The target audience of this document includes network equipment vendors, network operators, and standards development organizations. This document can be used as an index to some of the main OAM tools defined in the IETF. At the end of the document, a list of the OAM toolsets and a list of the OAM functions are presented as a summary. In situ Operations, Administration, and Maintenance (IOAM) collects operational and telemetry information in the packet while the packet traverses a path between two points in the network. This document provides a framework for IOAM deployment and provides IOAM deployment considerations and guidance. The One-Way Active Measurement Protocol (OWAMP) measures unidirectional characteristics such as one-way delay and one-way loss. High-precision measurement of these one-way IP performance metrics became possible with wider availability of good time sources (such as GPS and CDMA). OWAMP enables the interoperability of these measurements. [STANDARDS-TRACK] Huawei Technologies China Mobile Orange Telefonica Juniper Networks, Inc. This document describes a framework for Computing-Aware Traffic Steering (CATS). Particularly, the document identifies a set of CATS components, describes their interactions, and exemplifies the workflow of the control and data planes. Everything OPS Cisco Old Dog Consulting Nokia Blue Fern Consulting ZTE New Protocols or Protocol Extensions are best designed with due consideration of the functionality needed to operate and manage them. Retrofitting operations and management considerations is suboptimal. The purpose of this document is to provide guidance to authors and reviewers on what operational and management aspects should be addressed when defining New Protocols or Protocol Extensions. This document obsoletes RFC 5706, replacing it completely and updating it with new operational and management techniques and mechanisms. It also introduces a requirement to include an "Operational Considerations" section in new RFCs in the IETF Stream. China Mobile Huawei Technologies Telefonica Qualcomm Europe, Inc. Huawei Technologies Computing-Aware Traffic Steering (CATS) is a traffic engineering approach that optimizes the steering of traffic to a given service instance by considering the dynamic nature of computing and network resources. In order to consider the computing and network resources, a system needs to share information (metrics) that describes the state of the resources. Metrics from network domain have been in use in network systems for a long time. This document defines a set of metrics from the computing domain used for CATS. Informative References China Mobile Telefonica Huawei Technologies China Unicom Alibaba Group Distributed computing is a computing pattern that service providers can follow and use to achieve better service response time and optimized energy consumption. In such a distributed computing environment, compute intensive and delay sensitive services can be improved by utilizing computing resources hosted in various computing facilities. Ideally, compute services are balanced across servers and network resources to enable higher throughput and lower response time. To achieve this, the choice of server and network resources should consider metrics that are oriented towards compute capabilities and resources instead of simply dispatching the service requests in a static way or optimizing solely on connectivity metrics. The process of selecting servers or service instance locations, and of directing traffic to them on chosen network resources is called "Computing-Aware Traffic Steering" (CATS). This document provides the problem statement and the typical scenarios for CATS, which shows the necessity of considering more factors when steering traffic to the appropriate computing resource to better meet the customer's expectations. Old Dog Consulting This document describes the principles of traffic engineering (TE) in the Internet. The document is intended to promote better understanding of the issues surrounding traffic engineering in IP networks and the networks that support IP networking, and to provide a common basis for the development of traffic engineering capabilities for the Internet. The principles, architectures, and methodologies for performance evaluation and performance optimization of operational networks are also discussed. This work was first published as RFC 3272 in May 2002. This document obsoletes RFC 3272 by making a complete update to bring the text in line with best current practices for Internet traffic engineering and to include references to the latest relevant work in the IETF. This document describes the format of a set of control messages used in ICMPv6 (Internet Control Message Protocol). ICMPv6 is the Internet Control Message Protocol for Internet Protocol version 6 (IPv6). [STANDARDS-TRACK] This document describes a protocol intended to detect faults in the bidirectional path between two forwarding engines, including interfaces, data link(s), and to the extent possible the forwarding engines themselves, with potentially very low latency. It operates independently of media, data protocols, and routing protocols. [STANDARDS-TRACK] This document describes the use of the Bidirectional Forwarding Detection (BFD) protocol over IPv4 and IPv6 for single IP hops. [STANDARDS-TRACK] Many service provider service level agreements (SLAs) depend on the ability to measure and monitor performance metrics for packet loss and one-way and two-way delay, as well as related metrics such as delay variation and channel throughput. This measurement capability also provides operators with greater visibility into the performance characteristics of their networks, thereby facilitating planning, troubleshooting, and network performance evaluation. This document specifies protocol mechanisms to enable the efficient and accurate measurement of these performance metrics in MPLS networks. [STANDARDS-TRACK] The One-way Active Measurement Protocol (OWAMP), specified in RFC 4656, provides a common protocol for measuring one-way metrics between network devices. OWAMP can be used bi-directionally to measure one-way metrics in both directions between two network elements. However, it does not accommodate round-trip or two-way measurements. This memo specifies a Two-Way Active Measurement Protocol (TWAMP), based on the OWAMP, that adds two-way or round-trip measurement capabilities. The TWAMP measurement architecture is usually comprised of two hosts with specific roles, and this allows for some protocol simplifications, making it an attractive alternative in some circumstances. [STANDARDS-TRACK]

Contributors ZTE Corporation

huang.guangping@zte.com.cn

ZTE Corporation

huang.cheng13@zte.com.cn

ZTE Corporation

duan.wei1@zte.com.cn