]> ETRI

byyun@etri.re.kr

wooriNet

young@woori-net.com

CCAMP Working Group performance management streaming measurement This document provides a YANG data model of performance management streaming from network equipment to clients. It defines PM measurement methods, event notifications, generic PM parameters and streaming subscriptions. Additionally, it includes a YANG module for PM interval capabilities discovery, enabling clients to understand supported sampling and measurement intervals before configuring PM measurements. About This Document Status information for this document may be found at . Source for this draft and an issue tracker can be found at .

Introduction With the rise of AI-driven applications, network digital twins, and increasingly dynamic network environments, there is growing demand for performance management (PM) streaming capabilities. PM streaming enables proactive issue detection, allowing network operators to identify and address potential problems before they affect service. It also helps optimize resource allocation, ensuring efficient use of bandwidth and other network resources. provides a foundational framework for managing transport network elements, addressing requirements, parameters, and measurement methods for performance management. However, does not define YANG data models or specific protocols needed for PM streaming, which are essential for modern network management. To support PM streaming, various IETF documents and protocols , , can be utilized. This document provides a YANG data model for PM streaming in network equipment based on , demonstrating how to subscribe to the YANG model using the IETF push model.

PM Streaming

Streaming Network Architecture

PM streaming is a real-time method for measuring and transmitting data to monitor the performance and health of network devices and systems. It provides valuable insights into key metrics like errored seconds (ES), latency, and packet loss, helping to optimize networks, detect anomalies, and manage faults proactively. Unlike traditional periodic data collection, PM streaming delivers continuous updates, enabling faster, more responsive network adjustments. Using telemetry protocols like YANG Push, PM streaming allows for more frequent and detailed performance monitoring. By integrating this data into AI-driven analytics, it supports preemptive interventions, enhancing overall network reliability. Additionally, it keeps digital twins synchronized with the physical network, offering real-time insights for predictive maintenance, planning, and optimization. The procedures for Performance Management streaming between a network node and clients such as operations system (OS), AI applications, network digital twins (NDT) involve continuous measurement of performance metrics on PM parameters using three methods: counts (tracking event occurrences), snapshot (instantaneous metric values), and tidemarks (extreme values over a period). Clients can initiate the process by sending a subscription request specifying the metrics, measurement methods, intervals, and filtering criteria. Once the node confirms the subscription, it collects and aggregates PM data based on the requested metrics and intervals. Notifications with PM data, including timestamps, metrics, and measurement methods, are sent to clients at each interval via protocols like NETCONF or RESTCONF. Clients then process the data, using it for real-time monitoring, historical analysis, or triggering alerts based on thresholds. They can also manage subscriptions by modifying parameters or suspending them as needed.

PM parameters

Types Performance monitoring (PM) in networks encompasses a wide variety of parameters that reflect operational health, service quality, reliability, and environmental conditions. These parameters are used across many technologies, network layers, and functional domains to enable fault management, SLA compliance, trend analysis, predictive maintenance, and operational optimization. PM parameter types include but are not limited to:

• Classical transport and packet layer metrics: such as errored seconds (ES), severely errored seconds (SES), unavailable seconds (UAS), background block errors (BBE), background block counts (BBC), delay, jitter, and packet loss, as defined in standards like , and others.
• Layer-specific metrics:
  • Physical layer: optical power levels, laser bias current, loss of signal
  • Data link layer: Ethernet frame errors, FCS errors
  • Network layer: dropped packets, route flaps
  • Transport/Service layers: MPLS LSP statistics, OTN TCM/BIP counters
• Network environment parameters: including temperature, humidity, fan speed, voltage, and airflow. These are essential for equipment safety, energy management, and predictive failure analysis.
• Energy and sustainability metrics: such as power consumption, energy efficiency indicators, and cooling utilization, aligned with emerging sustainability standards and operational efficiency goals.
• Security and integrity parameters: such as pointer justification events (PJE), synchronization loss, or intrusion anomaly flags.
• Application-aware or SLA metrics: such as service availability, throughput consistency, and application-layer latency.
• Mobile network-specific metrics: including radio link failures, handover success/failure rates, RRC connection setup time, PDCP discard rate, and throughput per bearer. These metrics are critical for monitoring the performance of RAN, core, and edge network components in 4G/5G mobile environments.

These parameters may be grouped flexibly within the YANG model using parameter profiles that reflect shared characteristics, purpose, or applicable network domains. The architecture supports extension through identity-based typing to accommodate future parameter definitions introduced by standard bodies like ITU-T, IEEE, IETF, MEF, and TM Forum.

Profiles

Parameter Profile Subtree

The YANG model defines the concept of a parameter profile to logically group performance parameters that are commonly measured together for a specific operational purpose. Each parameter profile is represented as a list entry keyed by a name of type profile-names (a string conforming to the format defined in the model). These profiles serve as named collections of performance parameters and are intended to facilitate streamlined configuration, management, and reporting of measurement data across network elements and management systems. The use of parameter profiles improves operational efficiency by allowing operators, applications, and controllers to activate or reference a coherent set of parameters using a single profile identifier. For example, the itu-transport-maintenance-15min profile may include parameters such as errored seconds (ES), severely errored seconds (SES), and unavailable seconds (UAS), which are typically monitored together for network maintenance and fault detection purposes. Similarly, the ietf-qos-ip-24hr profile may include delay, jitter, and loss parameters used in service quality reporting. Parameter profiles support role-based access control, operational alignment, and measurement policy abstraction, enabling network operators and analytics systems to tailor data collection and reporting according to the needs of different users and services. The profile abstraction also aligns with , which identifies multiple classes of performance monitoring (e.g., maintenance, service-level, and compliance monitoring), each requiring specific sets of performance parameters. By modeling profiles as list entries keyed by a structured name (profile-names type), the YANG design ensures extensibility and vendor interoperability, allowing future profiles to be defined without changes to the core data structures. This approach promotes consistent configuration and integration across multi-vendor environments and supports dynamic service assurance use cases where parameter sets may vary by service type, SLA, or operational context.

Naming Parameter profiles are named to reflect their operational purpose, origin, applicable network domain, and, optionally, the primary measurement interval. This naming structure supports clarity, modularity, and automation across diverse network and service layers. The naming follows this format: <source>-<network>-<purpose>[-<characteristic>] Where:

• <source>: Standards body or organization
  • Examples: itu, ieee, ietf, vendorX
• <network>: Network domain or layer
  • Examples: transport, access, core, ip, mpls, ethernet, otn, wdm, flexo
• <purpose>: Intended use or function
  • Examples: maintenance, qos, availability, sla, compliance, analytics
• <characteristic> (optional): Optional qualifying information
  • Examples: 15min, 24hr, high-priority

Examples:

• itu-transport-maintenance-15min
• itu-transport-qos-24hr
• ieee-access-availability
• ietf-ip-qos-24hr
• vendorx-otn-sla

'transport' means that the profile applies to multiple technologies (e.g., OTN, MPLS-TP, Transport Ethernet, etc.). The 15-minute interval provides granular, real-time monitoring, allowing network operators to quickly detect and address short-term issues such as spikes in latency or packet loss. It is particularly useful for ensuring compliance with Service-Level Agreements (SLAs) and for managing highly dynamic networks where rapid changes can occur. In contrast, the 24-hour interval is used for long-term performance monitoring and trend analysis, helping operators understand overall network health, detect slow-developing issues, and plan for future capacity needs. This longer interval offers a broader view of the network's performance over a full day, making it ideal for strategic planning and infrastructure maintenance. Together, these intervals enable both immediate responses to network conditions and long-term network optimization.

Transport Common PM Parameters Metric value of PM parameters is measured for maintenance and QoS monitoring over networks. Quality of Service (QoS) parameters are designed to assess the network's long-term ability to consistently deliver agreed-upon service quality to customers. They primarily verify performance against contractual obligations defined in service-level agreements (SLAs) over longer intervals (24 hours, monthly periods). By simultaneously measuring both directions of a bidirectional connection, QoS parameters provide a holistic view of the sustained quality experienced by users, ensuring stability and predictability. Maintenance parameters focus on short-term monitoring and detailed analysis for operational reliability. Maintenance parameters, over intervals such as 15 minutes or 24 hours, facilitate swift responses to intermittent faults, bursts of errors, and subtle performance changes. Maintenance parameters typically involve unidirectional analysis, where each direction of transmission is monitored independently. This unidirectional approach helps network operators precisely pinpoint faults, troubleshoot intermittent issues, and perform preventive maintenance effectively. Key PM parameters focused on circuit networks such as OTN are listed as follows. Additional parameters will be needed for packet networks.

• ES Errored Seconds
• SES Severely Errored Seconds
• BBE Background Block Errors
• BBC Background Block Count
• UAS Unavailable Seconds
• SEP Severely Errored Period
• PJE Pointer Justification Events

According to the types of the measurement methods, purposes, and time intervals, different parameters are used. The following three transport profiles provide comprehensive coverage for maintenance and QoS monitoring across different temporal resolutions:

itu-transport-maintenance-15min Profile The itu-transport-maintenance-15min profile is designed for short-term operational monitoring and rapid fault detection. This profile utilizes all three measurement methods (counts, snapshot, and tidemarks) over 15-minute intervals to provide granular visibility into network performance. The profile includes core maintenance parameters: ES, SES, BBE, BBC, and UAS. The 15-minute interval enables operators to quickly detect and respond to performance degradation, making it ideal for proactive maintenance and immediate troubleshooting. The combination of counts (for cumulative event tracking), snapshot (for instantaneous state capture), and tidemarks (for extreme value monitoring) provides a comprehensive view of network behavior within each measurement period. This profile is particularly valuable for network operations centers (NOCs) requiring real-time visibility into transport network health and for automated systems that need to trigger immediate responses to performance anomalies.

itu-transport-maintenance-24hr Profile The itu-transport-maintenance-24hr profile extends the maintenance monitoring capabilities to longer-term analysis and trend identification. Similar to the 15-minute profile, it employs all three measurement methods (counts, snapshot, and tidemarks) but over 24-hour intervals. The parameter set includes ES, SES, BBE, BBC, UAS, and additionally PJE. The 24-hour measurement interval allows for comprehensive daily performance analysis, enabling operators to identify patterns, trends, and recurring issues that may not be apparent in shorter intervals. The inclusion of PJE provides additional insight into synchronization-related performance issues that are more relevant over longer observation periods. This profile supports strategic maintenance planning, capacity management, and historical performance analysis. It complements the 15-minute profile by providing the broader context needed for long-term network optimization and preventive maintenance strategies.

itu-transport-qos-24hr Profile The itu-transport-qos-24hr profile is specifically designed for Quality of Service monitoring and SLA compliance verification. Unlike the maintenance profiles, this profile focuses exclusively on the counts measurement method over 24-hour intervals, emphasizing sustained performance assessment rather than detailed operational monitoring. The parameter set includes ES, SES, BBE, BBC, SEP, and UAS. The QoS profile's exclusive use of counts measurement method reflects its focus on cumulative performance over extended periods, which is essential for SLA compliance and service quality reporting. The 24-hour interval aligns with typical SLA measurement periods and provides the statistical basis for service quality assessments. The inclusion of SEP is particularly relevant for QoS monitoring as it represents sustained periods of degraded performance that directly impact service quality commitments.

Profile Relationships and Operational Integration These three profiles work together to provide a comprehensive performance monitoring framework. The itu-transport-maintenance-15min profile serves as the primary operational tool for immediate network management, while the itu-transport-maintenance-24hr profile provides the analytical foundation for strategic planning and trend analysis. The itu-transport-qos-24hr profile ensures that service quality commitments are met and provides the data necessary for SLA reporting and customer assurance. The hierarchical relationship between these profiles enables operators to correlate short-term operational events with long-term performance trends and service quality metrics. For example, a spike in ES detected by the 15-minute maintenance profile can be correlated with daily trends from the 24-hour maintenance profile and assessed against SLA thresholds defined in the QoS profile. This integrated approach supports both reactive troubleshooting and proactive network optimization while ensuring compliance with service quality commitments.

Periodic Measurement

Measurement Timing

Sampling and Measurement Intervals Subtree

Measurement timing parameters are key components of network performance management, offering standardized definitions for the time-related aspects of sampling, measuring, and reporting performance data. These parameters apply to the three main measurement methods for network equipment: counts, snapshot, and tidemarks. They include the sampling interval, measurement interval, and uniform time, all of which support consistent, accurate, and systematic performance monitoring and management. Sampling interval defines the period at which network performance data is collected at consistent, predetermined time points. It ensures the continuous and timely capture of performance metrics, enabling accurate assessments of network conditions. Measurement interval specifies the duration over which sampled performance data is aggregated or statistically processed. It helps manage large volumes of data by summarizing it into meaningful indicators for analysis, anomaly detection, and resource management. Uniform time is a fixed, predefined point within each measurement interval at which a snapshot measurement is taken. It enables a consistent and instantaneous view of network performance across intervals, without requiring data aggregation. This approach facilitates quick diagnostics and synchronization across monitoring systems.

Use Cases The hierarchical design of the ietf-pm-measurements YANG module, wherein a performance parameter supports multiple sampling intervals and each sampling interval may be associated with multiple measurement intervals and methods, is motivated by a set of operationally validated use cases as shown in the following table. In these use cases, a single logical client-whether a human operator, network application, or analytics engine-requires simultaneous access to multiple views of the same performance parameter, differentiated by temporal resolution and analysis objective. In network operations centers (NOCs), an operator may require performance monitoring based on high-frequency samples (e.g., 1-second sampling with 15-minute measurement intervals), while concurrently assessing longer-term service-level agreement (SLA) compliance through longer measurement windows (e.g., 24-hour aggregation). It aligns with the performance monitoring functions and applications on performance monitoring parameters such as errored seconds (ES), etc. described in . A single dashboard or planning tool may correlate short-term utilization spikes with long-term trends using different sampling and measurement combinations on the same parameter. Network Performance Monitor for Critical Link Stability monitor tracks latency using the counts method at multiple time resolutions to address immediate service degradation and long-term path reliability. A single monitoring platform benefits from consistent sampling while leveraging different measurement intervals to inform short-term action and long-term optimization. An AI/ML analytics system may ingest the same parameter, such as ES at different granularities for multiple purposes: high-resolution data for anomaly detection, medium-resolution tidemarks for real-time model updates, and low-resolution tidemarks for daily or monthly training. All of these operations may be performed within the scope of a single analytics application. A digital twin platform continuously mirrors the real-time behavior of a physical network using packet delay variation (pdv). To accurately represent dynamic conditions, the system uses very fine-grained sampling with multiple count-based measurement intervals to feed simulation models and predictive engines in real time. Measurement data is ingested at short intervals (e.g., 1 minute) to maintain near-instantaneous synchronization with the physical network's current performance, supporting accurate digital mirroring. A slightly longer interval (e.g., 5 minutes) allows the twin to detect and buffer small fluctuations, supporting feedback loops that smooth reactive behavior and adapt to transient changes. Long-term intervals (e.g., 1 hour) feed higher-order analytics and AI-based learning models that detect drift, optimize resource allocation, and improve future simulation fidelity. In all these cases, the hierarchical list structure-where parameters contain multiple sampling intervals, and each sampling interval defines one or more measurement intervals and methods-supports operational flexibility, avoids configuration duplication, and enables fine-grained control of measurement strategies.

Use cases of sampling and measurement intervals

Measurement Methods

Measurement Methods Subtree

The measurement methods defined based on establish a focused and efficient framework for network performance monitoring by specifying three core methods: counts, snapshot, and tidemarks. This intentional limitation supports key objectives such as implementation simplicity, operational efficiency, and cross-vendor interoperability. It emphasizes real-time network monitoring, favoring instantaneous or interval-based metrics over complex statistical calculations. Counts and snapshot methods provide immediate operational data without incurring the processing overhead associated with metrics like averages and variances etc. These statistical measures require significant aggregation logic, which can vary across implementations and devices. By keeping computation within network elements minimal, the approach reduces both processing and memory overhead, maintaining lightweight implementations. It establishes a clear separation between raw data collection (handled by network elements) and deeper analysis (delegated to external management systems). This separation not only simplifies device requirements but also enables more consistent and flexible analytics in centralized systems, which are better equipped to apply standardized analytical frameworks. Limiting measurement methods also contributes to energy efficiency by reducing the operational burden on Network Elements (NEs), while offloading data analysis to external management applications. Despite the simplicity, the selected measurement methods offer sufficient expressiveness to support comprehensive performance monitoring without excessive resource use. So, they are specifically optimized for Southbound Interface (SBI) between Physical Network Controllers (PNCs) and NEs to ensure as follows:

• Lightweight to implement
• Consistently supported across vendors
• Efficient for transport and storage in network management systems

The measurement methods are applicable to a wide range of monitored objects, including both network topology elements (e.g., links, tunnels) and physical equipment parameters(e.g., temperature, voltage).

Counts Counts measurement in network performance monitoring tracks the cumulative occurrences of specific events over a defined measurement interval, such as 15 minutes or 24 hours. This method captures how frequently certain network activities, like errors or transmission issues, occur, providing a historical view of recurring problems. Counts reset at the end of each interval, ensuring that every period starts with a fresh count for accurate monitoring. The primary purpose of counts is to identify trends and patterns in network behavior over time, helping operators detect anomalies or areas where issues frequently arise. This type of measurement is particularly useful for long-term analysis, enabling preventive maintenance and optimizing network performance. Unlike instantaneous measurements, counts focus on aggregation over time, making it easier to understand the persistence or recurrence of faults. The data gathered through counts helps in fault management and planning by highlighting repeated errors, congestion, or performance degradation that may affect service delivery. As a result, counts provide network operators with actionable insights for troubleshooting and capacity planning, ensuring smooth operation and reliability across the network.

Snapshot Snapshot is an instantaneous measurement taken at a specific point in time. It captures the instantaneous value of specific performance parameters at a regular, predefined point (uniform time) within each time interval. Snapshot provides a "momentary view" of network conditions, allowing operators to observe the network's status at specific intervals. The data from these uniform-time snapshots is then aggregated and analyzed to understand the immediate state across the entire network. By taking snapshots simultaneously across all network elements, operators can correlate data between different parts of the transport network. Snapshots are collected at pre-determined uniform times within fixed measurement intervals. The uniform time and fixed intervals can be configured based on the needs of the network.

Tidemarks Tidemarks measurements record the maximum (high tidemarks) and minimum (low tidemarks) values that a performance parameter reaches during a specified measurement interval. This approach captures the extreme values and performance fluctuations, highlighting the best and worst conditions that occur within the monitoring period. Tidemarks measurements provide deeper insights by capturing performance spikes or drops that may go unnoticed in average or cumulative data, enabling precise troubleshooting of intermittent or extreme conditions. For instance, while the average error rate over a period may appear acceptable, a high tidemarks could reveal intermittent spikes in errors that require attention. Conversely, a low tidemarks may expose periods of severely degraded signal quality or throughput.

Thresholding

Periodic Thresholding Periodic threshold events are triggered when the counts or gauge value reaches a pre-defined threshold during periodic measurements including counts, snapshot, and tidemarks for performance parameters. The counts measurement has two types of threshold reporting methods: transient and standing condition methods. The transient condition method treats each measurement period separately. As soon as a threshold is reached or crossed in a measurement interval for a given performance measurement, a threshold report (TR) is generated. The standing condition method is optional. The standing condition is raised, and a TR (Threshold Report) is generated, when the set threshold is reached or crossed. The standing condition is cleared, and a reset threshold report (RTR) is generated at the end of the period when the current value is below or equal to the reset threshold, provided that there was no unavailable time during that period. For gauge measurements ("snapshot" and "tidemarks"), an overflow condition is determined and an out-of-range report is generated as soon as the gauge value reaches or crosses the high threshold. An underflow condition is determined and an out-of-range report is generated as soon as the gauge value is at or below the low threshold.

Non-Periodic Thresholding Non-periodic threshold events are triggered regardless of the measurement methods (counts, snapshot, or tidemarks). The following parameters are used for non-periodic events.

• BUT (Begin Unavailable Time): The event marking the start of a period when a network element or connection is unavailable.
• EUT (End Unavailable Time): The event marking the end of a period when a network element or connection was unavailable.
• CSES (Consecutive Severely Errored Seconds): A sequence of severely errored seconds (SES) detected consecutively within a specified time interval. The reporting metrics include BUT, EUT, and the count of errors during that period.

Subscriptions Clients can receive a streaming of values of the PM parameters (PM data) by the measurement methods of counts, snapshot, and tidemarks with various time intervals, after subscribing to them in equipment. The streaming data measured on the PM parameters are used for maintenance and Quality of Service (QoS) monitoring in networks. Below are practical use cases demonstrating different subscription scenarios. These examples illustrate periodic and non-periodic subscriptions, including notifications triggered by threshold breaches. Each example aligns with IETF YANG Push protocols, showcasing how network elements generate and stream performance data based on subscription parameters.

Periodic Events The YANG Push subscription model, as defined in , enables clients to subscribe to periodic performance measurement data from network elements. This model supports dynamic subscription establishment, modification, and termination for real-time streaming of PM data. Clients can specify subscription parameters including the target datastore (operational), encoding format (XML/JSON), and filtering criteria to receive only relevant performance metrics. The subscription mechanism allows for configurable update periods, enabling both high-frequency monitoring and long-term trend analysis (e.g., 24-hour intervals). Network elements generate periodic event notifications containing the requested PM data, which clients can process for real-time monitoring, historical analysis, or triggering automated responses based on performance thresholds. shows a subscription request for the ES parameter in the itu-transport-maintenance-15min profile. It requests counts measurement data sampled every second and aggregated over a 15-minute interval. The reporting period is set to 900 seconds, so a notification is sent at the end of each measurement interval.

Periodic Event Subscription Example YANG-PUSH encode-xml operational /pm-meas:pm-periodic-measurement/ parameter-profile[name='itu-transport-maintenance-15min']/ pm-parameter[name='es']/ sampling-interval[id='1s']/ measurement-interval[id='15min']/ measurement-methods/counts/measurement-value 900 2024-07-01T00:00:00Z ]]>

This XML shows a notification for the ES parameter in the itu-transport-maintenance-15min profile. It reports the counts measurement value sampled every second and aggregated over a 15-minute interval. The measured value (10) represents the total errored seconds in that period.

Periodic Event Notification Example 2024-07-01T00:15:00Z itu-transport-maintenance-15min es 1s 1 second 15min 15 minute 10 ]]>

Threshold Events Threshold event subscriptions enable clients to receive immediate notifications when performance metrics cross predefined thresholds, providing proactive monitoring capabilities. This subscription type, based on YANG Push for datastore change notifications, allows clients to define threshold conditions.

Periodic Threshold Events

Periodic Threshold Events Subtree

When a performance parameter exceeds or falls below configured thresholds for the periodic measurement methods of counts, snapshot, and tidemarks, the network element generates event-driven notifications containing the threshold crossing event type and occurrence time (parameter values at the time of the event can be read from the operational datastore if needed). This mechanism supports four types of threshold events: count-transient-event for immediate threshold crossings, count-standing-event for persistent threshold violations, snapshot-event for instantaneous value threshold crossings, and tidemark-event for extreme value threshold crossings. These events enable rapid response to network performance degradation and automated fault management. The threshold event subscription complements periodic subscriptions by providing real-time alerts for critical performance issues that require immediate attention. shows an example of the NETCONF request to subscribe to all pm-threshold-events notifications in the ietf-pm-measurements model.

Threshold Event Subscription Example YANG-PUSH encode-xml operational /pm-meas:pm-threshold-events 1 2024-07-01T00:00:00Z ]]>

reports a high-oor-event threshold crossing for the snapshot measurement of the ES parameter in the itu-transport-maintenance-15min profile, with 1-second sampling and 15-minute measurement interval. It shows the event type, occurrence, and timestamp as defined in the YANG model.

Threshold Event Notification Example 2024-07-01T00:05:23Z itu-transport-maintenance-15min es 1s 1 second 15min 15 minute High-OOR-event true 2024-07-01T00:05:23Z ]]>

Non-Periodic Threshold Events

Non Periodic Threshold Events Subtree

Non-periodic threshold event subscriptions provide immediate notifications for critical network availability and error conditions that occur independently of regular measurement intervals. These subscriptions monitor for specific events such as BUT, EUT, and CSES that indicate significant network performance degradation or service interruptions. When these events occur, the network element generates immediate notifications containing event details, timing information, and duration data. This subscription type enables proactive network management by providing real-time awareness of critical network conditions that require immediate operator attention or automated intervention. Non-periodic threshold events complement periodic monitoring by capturing exceptional conditions that may not be detected through regular interval-based measurements.

YANG Data Model The YANG module for PM measurements is defined below: Editor: Bin Yeong Yoon "; description "This YANG module defines a comprehensive data model for performance management (PM) streaming from network equipment to clients, based on ITU-T G.7710. It supports real-time streaming of performance measurements using three core methods: counts (cumulative events), snapshot (instantaneous values), and tidemarks (extreme values). The module enables proactive network monitoring through configurable sampling and measurement intervals, supporting both high-frequency real-time monitoring and long-term trend analysis. It provides threshold event notifications for both periodic measurements and non-periodic events (BUT, EUT, CSES). The design supports AI-driven applications, network digital twins, and dynamic network environments by enabling multiple simultaneous views of the same performance parameter with different temporal resolutions. This hierarchical structure allows operators, analytics systems, and digital twin platforms to access performance data at appropriate granularities while maintaining operational efficiency and cross-vendor interoperability. The module integrates with IETF YANG Push protocols for subscription-based streaming, enabling clients to receive continuous performance data and threshold event notifications for real-time monitoring, historical analysis, and automated network management."; revision 2026-02-06 { description "Updated transient-condition-config to only support transient-threshold, removing low-threshold support. Transient thresholds do not report low OOR (underflow) conditions, unlike snapshot and tidemarks measurements which support both high and low thresholds."; reference "RFC XXXX: A YANG Data Model of Performance Management Streaming"; } /* * TYPEDEFs */ typedef profile-names { type string { pattern '[a-zA-Z][a-zA-Z0-9_-]*-[a-zA-Z][a-zA-Z0-9_-]*-[a-zA-Z][a-zA-Z0-9_-]*(-[a-zA-Z][a-zA-Z0-9_-]*)?'; } description "Parameter profile name following the format: --[-] Where: - : Standards body or organization (e.g., itu, ieee, ietf) - : Network domain or layer (e.g., transport, access, core) - : Intended use or function (e.g., maintenance, qos, availability) - : Optional qualifying information (e.g., 15min, 24hr, high-priority) Examples: - itu-transport-maintenance-15min - itu-transport-maintenance-24hr - itu-transport-qos-24hr - ieee-access-availability"; } typedef time-interval-unit { type enumeration { enum millisecond { description "Time interval in milliseconds."; } enum second { description "Time interval in seconds."; } enum minute { description "Time interval in minutes."; } enum hour { description "Time interval in hours."; } } description "Units for expressing time intervals."; } /* * IDENTITIES */ identity periodic-events { description "Base identity for periodic event notifications."; } identity counts-transient { base periodic-events; description "Notification for transient threshold events in counts measurements."; } identity counts-standing { base periodic-events; description "Notification for standing threshold events in counts measurements."; } identity snapshot { base periodic-events; description "Notification for snapshot measurement threshold events."; } identity tidemarks { base periodic-events; description "Notification for tidemarks measurement threshold events."; } identity non-periodic-events { description "Base identity for non-periodic event notifications."; } identity but { base non-periodic-events; description "Notification for Begin Unavailable Time (BUT) events."; } identity eut { base non-periodic-events; description "Notification for End Unavailable Time (EUT) events."; } identity cses { base non-periodic-events; description "Notification for Consecutive Severely Errored Seconds (CSES) events."; } /* * COMMON GROUPINGS */ grouping threshold-config { description "Common threshold configuration for snapshot and tidemarks measurement types (high and low thresholds)."; leaf high-threshold { type uint32; description "High threshold that triggers alerts when exceeded."; } leaf low-threshold { type uint32; description "Low threshold that triggers alerts when performance falls below acceptable levels."; } } grouping transient-threshold-config { description "Threshold configuration for transient conditions. Transient thresholds only support high threshold crossings and report immediately when the count value reaches or crosses the configured transient threshold value. Transient thresholds do not support low threshold (underflow) conditions, unlike snapshot and tidemarks measurements which support both high and low thresholds."; leaf transient-threshold { type uint32; description "Transient threshold that triggers alerts when exceeded. Transient thresholds report immediately when the count value reaches or crosses this threshold value."; } } grouping event-state-info { description "Common event state information for all event types."; leaf event-occurred { type boolean; description "Indicates whether a threshold crossing or performance event has occurred."; } leaf event-time { type yang:date-and-time; description "Precise timestamp of when the event occurred."; } } grouping oor-event-type { description "Common out-of-range event type definition."; leaf event-type { type enumeration { enum High-OOR-event { description "High OOR threshold exceeded."; } enum Low-OOR-event { description "Low OOR threshold crossed."; } } description "Specifies whether the high or low OOR threshold was crossed."; } } grouping triggered-oor-event-info { description "Combined threshold event type and event information."; uses oor-event-type; uses event-state-info; } grouping count-transient-event-type { description "Transient threshold event type definition for counts measurements. Transient thresholds report immediately when the count value reaches or crosses a configured threshold value."; leaf event-type { type enumeration { enum Threshold-Crossed-Event { description "Threshold crossing event generated when count value reaches or crosses the configured threshold value."; } } description "Specifies that a threshold crossing event occurred."; } } grouping triggered-count-transient-event-info { description "Combined transient threshold event type and event information for counts measurements. Transient thresholds are independent threshold mechanisms that report immediately when count values cross configured threshold values."; uses count-transient-event-type; uses event-state-info; } grouping time-interval-config { description "Common time interval configuration."; leaf interval-value { type uint32; description "Numeric value for the interval."; } leaf unit { type time-interval-unit; description "Time unit for the interval value."; } } /* * MEASUREMENT METHOD GROUPINGS */ grouping count-measurement-gr { description "Counts measurement for cumulative event tracking over a measurement interval. Supports transient and standing threshold reporting, as defined in G.7710."; container counts { description "Contains counts measurement values and configuration."; container transient-condition-config { description "Configuration for transient threshold conditions. Transient thresholds report immediately when the count value reaches or crosses the configured transient threshold value. Transient thresholds do not support low threshold (underflow) conditions, unlike snapshot and tidemarks measurements which support both high and low thresholds."; uses transient-threshold-config; } container standing-condition-config { description "Configuration for standing condition monitoring. When both thresholds are set, standing-threshold must be greater than or equal to reset-threshold (hysteresis)."; must "not(standing-threshold and reset-threshold) or " + "standing-threshold >= reset-threshold" { error-message "Standing threshold must be >= reset threshold."; } leaf standing-threshold { type uint32; description "Threshold value that triggers standing condition alerts."; } leaf reset-threshold { type uint32; description "Reset threshold value that clears standing conditions."; } } leaf measurement-value { type uint32; config false; description "Current cumulative count value for the measurement interval."; } } } grouping snapshot-measurement-gr { description "Snapshot measurements for instantaneous values at uniform time within each measurement interval. Supports high/low OOR threshold reporting, as defined in G.7710."; container snapshot { description "Contains snapshot measurement configuration and values."; container uniform-time-config { description "Configuration for uniform time intervals between snapshots."; leaf interval-value { type uint32; default 1; description "Numeric value for the sampling interval between snapshots."; } leaf unit { type time-interval-unit; description "Time unit for the snapshot sampling interval."; } } container threshold-config { description "Configuration for snapshot threshold monitoring."; uses threshold-config; } leaf measurement-value { type uint32; config false; description "Current instantaneous snapshot value."; } } } grouping tidemarks-measurement-gr { description "Tidemarks measurements for maximum and minimum values over the measurement interval. Supports high/low OOR threshold reporting, as defined in G.7710."; container tidemarks { description "Contains tidemarks measurement values and threshold configuration."; container threshold-config { description "Configuration for tidemarks threshold monitoring."; uses threshold-config; } leaf high-measurement-value { type uint32; config false; description "Current maximum value recorded during the measurement interval."; } leaf low-measurement-value { type uint32; config false; description "Current minimum value recorded during the measurement interval."; } } } grouping measurement-methods-gr { description "Grouping for the three core measurement methods (counts, snapshot, tidemarks)."; container measurement-methods { description "Container for different measurement methods."; uses count-measurement-gr; uses snapshot-measurement-gr; uses tidemarks-measurement-gr; } } /* * EVENT GROUPINGS */ grouping counts-transient-event-gr { description "Transient threshold events for counts measurements. Transient thresholds are independent threshold mechanisms that report immediately when count values cross configured threshold values."; container counts-transient { description "Contains information about transient threshold events for counts."; uses triggered-count-transient-event-info; } } grouping counts-standing-event-gr { description "Standing condition events for counts measurements."; container counts-standing { description "Contains information about standing threshold events for counts."; leaf event-type { type enumeration { enum Threshold-Report { description "Threshold Report (TR) generated when the count value reaches or exceeds the standing-threshold configured in standing-condition-config."; } enum Reset-Threshold-Report { description "Reset Threshold Report (RTR) generated at the end of the period when the count value is at or below the reset-threshold (G.7710 standing condition clear)."; } } description "Specifies the type of standing threshold event that occurred, as defined in G.7710. A Threshold-Report (TR) is generated when the measurement-value reaches or exceeds the standing-threshold. A Reset-Threshold-Report (RTR) is generated at the end of the period when the measurement-value is at or below the reset-threshold."; } uses event-state-info; } } grouping snapshot-events-gr { description "Threshold events for snapshot measurements."; container snapshot { description "Contains snapshot threshold event information."; uses triggered-oor-event-info; } } grouping tidemarks-events-gr { description "Threshold events for tidemarks measurements."; container tidemarks { description "Contains tidemarks threshold event information."; uses triggered-oor-event-info; } } /* * MEASUREMENT INTERVAL STRUCTURES */ grouping periodic-measurement-intervals { description "Hierarchical structure for periodic measurement timing and methods."; list sampling-interval { key "id"; description "List of sampling intervals defining data collection frequency."; leaf id { type string; description "Unique identifier for this sampling interval configuration."; } leaf interval-value { type uint32; default 1; description "Numeric value for the sampling interval."; } leaf unit { type time-interval-unit; default second; description "Time unit for the sampling interval value."; } list measurement-interval { key "id"; description "List of measurement intervals defining aggregation periods."; leaf id { type string; description "Unique identifier for this measurement interval configuration."; } leaf interval-value { type uint32; default 15; description "Numeric value for the measurement interval."; } leaf unit { type time-interval-unit; default minute; description "Time unit for the measurement interval value."; } uses measurement-methods-gr; } } } grouping non-periodic-events-gr { description "Grouping for non-periodic performance event parameters (BUT, EUT, CSES)."; container BUT-event { description "Begin Unavailable Time (BUT) event."; uses event-state-info; } container EUT-event { description "End Unavailable Time (EUT) event."; uses event-state-info; leaf duration { type uint32; units "seconds"; description "Total duration of unavailability in seconds."; } } container CSES-event { description "Consecutive Severely Errored Seconds (CSES) event."; leaf event-occurred { type boolean; description "Indicates whether a CSES event was generated."; } leaf start { type yang:date-and-time; description "Timestamp indicating when the CSES period began."; } leaf end { type yang:date-and-time; description "Timestamp indicating when the CSES period ended."; } leaf duration { type uint32; units "seconds"; description "Duration of the CSES period in seconds."; } leaf error-count { type uint32; description "Number of errors during the CSES period."; } } } grouping pm-periodic-measurement-gr { description "Hierarchical structure for periodic performance measurements."; list parameter-profile { key "name"; description "List of performance parameter profiles."; leaf name { type profile-names; description "Name of the parameter profile."; } list pm-parameter { key "name"; description "List of PM parameters within the parameter profile."; leaf name { type string; description "Name of the performance parameter being measured."; } uses periodic-measurement-intervals; } } } /* * MAIN CONTAINER */ container pm-periodic-measurement { description "Main container for periodic performance measurements."; uses pm-periodic-measurement-gr; } /* * NOTIFICATIONS */ notification pm-threshold-events { description "Notification for periodic threshold crossing events and non-periodic performance events (BUT, EUT, CSES)."; container periodic-events { description "Container for periodic threshold events."; list parameter-profile { key "name"; description "List of performance parameter profiles for event monitoring."; leaf name { type profile-names; description "Name of the parameter profile."; } list pm-parameter { key "name"; description "List of PM parameters within the parameter profile."; leaf name { type string; description "Name of the performance parameter being monitored."; } list sampling-interval { key "id"; description "List of sampling intervals for event monitoring."; leaf id { type string; description "Unique identifier for this sampling interval configuration."; } uses time-interval-config; list measurement-interval { key "id"; description "List of measurement intervals for event aggregation."; leaf id { type string; description "Unique identifier for this measurement interval configuration."; } uses time-interval-config; container event-types { description "Container for different threshold event types."; uses counts-transient-event-gr; uses counts-standing-event-gr; uses snapshot-events-gr; uses tidemarks-events-gr; } } } } } } container non-periodic-events { description "Container for non-periodic performance events (BUT, EUT, CSES)."; uses non-periodic-events-gr; } } } ]]>

YANG Data Trees

Tree of pm measurements module

PM Interval Capabilities The ietf-pm-interval-capabilities YANG module provides comprehensive capability discovery for interval configurations, enabling clients to understand the supported temporal resolutions before configuring PM measurements. This module is designed to work seamlessly with the ietf-pm-measurements module, supporting both real-time monitoring scenarios requiring high-frequency sampling and long-term trend analysis requiring extended measurement intervals. The advertisement of interval capabilities follows standard IETF procedures for capability discovery in network management systems, based on the principles outlined in for YANG module capability advertisement, for YANG library, and for YANG module advertisement.

Motivation The need for interval capabilities discovery arises from several critical factors that affect the configuration and operation of performance monitoring systems. Different types of performance parameters inherently require different sampling and measurement intervals based on their characteristics and measurement objectives. For example, error-based parameters such as errored seconds (ES) may require frequent sampling to capture transient events, while availability metrics might be adequately monitored with longer intervals. Similarly, latency measurements often need high-frequency sampling to detect microsecond-level variations, whereas throughput statistics can be effectively captured with less frequent sampling. Vendor dependencies represent another significant factor necessitating interval capabilities discovery. Network equipment manufacturers implement different hardware architectures, processing capabilities, and measurement engines, resulting in varying support for sampling and measurement intervals. Some vendors may support very fine-grained intervals (e.g., millisecond-level sampling) for high-precision applications, while others may be optimized for longer intervals suitable for operational monitoring. Additionally, different vendors may have different constraints on the relationship between sampling and measurement intervals, with some supporting only specific multiples or ranges. The complexity of modern network environments further underscores the importance of interval capabilities discovery. Multi-vendor networks require clients to adapt their monitoring strategies based on the specific capabilities of each network element. Without proper capability discovery, clients risk configuring unsupported intervals, leading to configuration failures, suboptimal monitoring, or even system instability. The interval capabilities framework addresses these challenges by providing a standardized mechanism for discovering and understanding the temporal resolution capabilities of network elements, enabling clients to make informed decisions about interval configuration and ensuring interoperability across diverse network environments. The interval capabilities module follows a hierarchical structure that mirrors the measurement configuration model, ensuring consistency between capability discovery and actual measurement configuration. The architecture consists of three levels: Parameter Profiles (collections of related performance parameters such as itu-transport-maintenance-15min), PM Parameters (individual performance parameters within profiles such as es, ses, bbe), and Interval Capabilities (sampling and measurement interval capabilities for each parameter). The module defines a critical relationship between sampling and measurement intervals: measurement intervals must be multiples of their corresponding sampling intervals. This constraint ensures that measurement aggregation periods align with data collection frequency, preventing configuration errors and ensuring accurate performance monitoring. For example, if a device supports a 5-second sampling interval, valid measurement intervals would be 5s, 10s, 15s, 30s, 60s, etc. This relationship is enforced through the hierarchical structure where measurement intervals are defined within their corresponding sampling intervals.

Capability Discovery and Configuration Workflow A NETCONF client can discover the sampling and measurement interval capabilities of a server by following the standard IETF procedures for capability and module discovery. This process involves multiple steps, beginning with the session establishment and extending to operational data retrieval. Upon initiating a session, the client receives the server's <hello> message as defined in . This message includes a list of capability URIs, indicating the supported YANG modules and protocol extensions. If the server includes entries for both ietf-pm-measurements and ietf-pm-interval-capabilities, the client infers that the server supports performance measurement with parameter-specific intervals, and also advertises its supported interval values. To confirm module support and retrieve metadata such as revision dates and feature availability, the client queries the YANG Library as defined in . This step allows the client to verify that both the measurement model and the interval capability model are implemented and discoverable. The client then queries the pm-interval-capabilities container, which is defined with config false and thus resides in the operational datastore per the Network Management Datastore Architecture (NMDA) described in . By querying this container, the client can retrieve a list of supported sampling and measurement intervals for each performance parameter and profile. The structure includes constraints such as minimum and maximum values, allowed time units, and granularity. This live runtime exposure of capability information follows the model described in for advertising telemetry and notification capabilities in operational state. Alternatively, the same interval capabilities may be published as static data files using the format defined in . This allows vendors or standards bodies to document the supported measurement intervals of a device or profile without requiring a live connection to the system. At this point, the client uses the retrieved interval capabilities to configure a performance measurement subscription using the pm-measurements model. This configuration is made in alignment with the update intervals supported by the server, ensuring compatibility and preventing errors such as period-unsupported, as outlined in . This end-to-end process ensures that performance measurement configurations are both valid and optimized for the server's capabilities, leveraging both static publication and runtime introspection using standardized models and procedures.

Interval Capabilities Example The following example demonstrates how a client can discover interval capabilities for the ES parameter in the itu-transport-maintenance-15min profile, specifically requesting 1-second sampling with 15-minute measurement intervals.

Interval Capabilities Discovery Example itu-transport-maintenance-15min es 1s 1 1 second 1 second 1 measurement-range 5 1440 minute 15 minute 5 ]]>

This example shows how a client can discover that the network element supports 1-second sampling with a flexible measurement interval range (5-1440 minutes) with 5-minute granularity. The response confirms unit support for seconds in sampling and minutes in measurement intervals, with a default recommendation of 15 minutes, enabling the client to choose any appropriate measurement duration within the supported range for their PM monitoring requirements.

YANG Data Model Editor: Bin Yeong Yoon "; description "This YANG module defines a data model for discovering and expressing performance management (PM) interval capabilities in network equipment. The module enables clients to discover what sampling and measurement intervals a server can support for different performance parameters within various parameter profiles. This module provides hierarchical interval capability discovery where measurement intervals must be multiples of their corresponding sampling intervals, and each parameter can have different interval capabilities within parameter profiles. This module is designed to work with ietf-pm-measurements for complete PM streaming solutions and supports both real-time monitoring and long-term trend analysis."; revision 2025-10-17 { description "Initial version."; reference "RFC XXXX: A YANG Data Model of Performance Management Streaming"; } /* * TYPEDEFs */ typedef interval-unit { type enumeration { enum millisecond { description "Time interval in milliseconds."; } enum second { description "Time interval in seconds."; } enum minute { description "Time interval in minutes."; } enum hour { description "Time interval in hours."; } enum day { description "Time interval in days."; } } description "Supported units for expressing time intervals."; } /* * IDENTITIES */ identity interval-capability-type { description "Base identity for different types of interval capabilities."; } identity sampling-interval-capability { base interval-capability-type; description "Capability for sampling intervals - how frequently data is collected."; } identity measurement-interval-capability { base interval-capability-type; description "Capability for measurement intervals - how long measurements are aggregated."; } /* * GROUPINGS */ grouping interval-constraints { description "Constraints for supported intervals including min/max values and supported units. This grouping defines the capability constraints for both sampling and measurement intervals, allowing devices to express their supported interval ranges, units, and granularity."; leaf min-value { type uint32; description "Minimum supported value for this interval type."; } leaf max-value { type uint32; description "Maximum supported value for this interval type."; } leaf-list units { type interval-unit; description "List of supported time units for this interval type."; } leaf default-value { type uint32; description "Default value recommended for this interval type."; } leaf default-unit { type interval-unit; description "Default unit recommended for this interval type."; } leaf granularity { type uint32; description "Granularity step for interval values. For example, if granularity is 5, then only values that are multiples of 5 are supported."; } } grouping parameter-interval-capabilities { description "Interval capabilities for a specific parameter within a profile, including hierarchical interval relationships where all interval information is contained within the sampling-interval structure."; leaf name { type string; description "Name of the performance parameter (e.g., es, ses, bbe)."; } container interval-relationships { description "Defines hierarchical relationships between sampling and measurement intervals for this parameter. Measurement intervals must be multiples of their corresponding sampling intervals."; list sampling-interval { key "id"; description "Maps sampling intervals to their compatible measurement intervals with complete capability information."; leaf id { type string; description "Unique identifier for this sampling interval capability."; } uses interval-constraints; list measurement-interval { key "id"; description "Detailed information for each compatible measurement interval within the sampling interval structure."; leaf id { type string; description "Unique identifier for this measurement interval capability."; } uses interval-constraints; } } } } grouping profile-parameter-capabilities { description "Capabilities for all parameters within a specific parameter profile. This grouping defines the complete capability information for a parameter profile, including the profile name and all performance parameters with their respective interval capabilities."; leaf name { type pm-meas:profile-names; description "Name of the parameter profile (e.g., itu-transport-maintenance-15min)."; } list pm-parameter { key "name"; description "List of parameters with their specific interval capabilities within this profile."; uses parameter-interval-capabilities; } } /* * MAIN CONTAINER */ container pm-interval-capabilities { description "Main container for hierarchical PM interval capabilities. This container provides comprehensive information about the sampling and measurement intervals that a server can support for different parameters within different parameter profiles."; config false; list parameter-profile { key "name"; description "List of parameter profiles with their parameter-specific interval capabilities. Each profile represents a collection of parameters that share common measurement requirements but may have different interval capabilities."; uses profile-parameter-capabilities; } } } ]]>

YANG Data Trees

Tree of pm interval capabilities module

Manageability Considerations <Add any manageability considerations>

Security Considerations <Add any security considerations>

IANA Considerations <Add any IANA considerations>

References Normative References This document defines a YANG data model and associated mechanisms enabling subscriber-specific subscriptions to a publisher's event streams. Applying these elements allows a subscriber to request and receive a continuous, customized feed of publisher-generated information. This document provides a Network Configuration Protocol (NETCONF) binding to the dynamic subscription capability of both subscribed notifications and YANG-Push. This document describes a mechanism that allows subscriber applications to request a continuous and customized stream of updates from a YANG datastore. Providing such visibility into updates enables new capabilities based on the remote mirroring and monitoring of configuration and operational state. There is a need to document data defined in YANG models at design time, implementation time, or when a live server is unavailable. This document specifies a standard file format for YANG instance data, which follows the syntax and semantics of existing YANG models and annotates it with metadata. This document describes a YANG library that provides information about the YANG modules, datastores, and datastore schemas used by a network management server. Simple caching mechanisms are provided to allow clients to minimize retrieval of this information. This version of the YANG library supports the Network Management Datastore Architecture (NMDA) by listing all datastores supported by a network management server and the schema that is used by each of these datastores. This document defines two YANG modules, "ietf-system-capabilities" and "ietf-notification-capabilities". The module "ietf-system-capabilities" provides a placeholder structure that can be used to discover YANG-related system capabilities for servers. The module can be used to report capability information from the server at runtime or at implementation time by making use of the YANG instance data file format. The module "ietf-notification-capabilities" augments "ietf-system-capabilities" to specify capabilities related to "Subscription to YANG Notifications for Datastore Updates" (RFC 8641). The Network Configuration Protocol (NETCONF) defined in this document provides mechanisms to install, manipulate, and delete the configuration of network devices. It uses an Extensible Markup Language (XML)-based data encoding for the configuration data as well as the protocol messages. The NETCONF protocol operations are realized as remote procedure calls (RPCs). This document obsoletes RFC 4741. [STANDARDS-TRACK] Datastores are a fundamental concept binding the data models written in the YANG data modeling language to network management protocols such as the Network Configuration Protocol (NETCONF) and RESTCONF. This document defines an architectural framework for datastores based on the experience gained with the initial simpler model, addressing requirements that were not well supported in the initial model. This document updates RFC 7950. Informative References ITU-T

Contributors KT

kwangkoog.lee@kt.com

SK Telecom

jongyoon.shin@sk.com

LGU+

sy.nam@lguplus.co.kr