Internet-Draft Incident Management March 2023
Feng, et al. Expires 14 September 2023 [Page]
Workgroup:
OPSAWG
Internet-Draft:
draft-feng-opsawg-incident-management-00
Published:
Intended Status:
Standards Track
Expires:
Authors:
C. Feng, Ed.
Huawei
T. Hu
CMCC
LM. Contreras
Telefonica I+D
Q. Wu
Huawei
C. Yu
Huawei

Incident Management for Network Services

Abstract

This document provides an architecture for the incident management system and related function interface requirements.

This document also defines a YANG module to support the incident lifecycle management. This YANG module is meant to provide a standard way to report, diagnose, and resolve incidents for the sake of enhanced network services.

Status of This Memo

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."

This Internet-Draft will expire on 14 September 2023.

Table of Contents

1. Introduction

Network performance management and fault management are used for monitoring and troubleshooting separately in networking infrastructures. Typically, metrics and alarms, transaction operations are monitored centrally and incident tickets are triggered accordingly. A YANG [RFC7950] data model for alarm management [RFC8632] defines a standard interface for alarm management.

A data model for Network and VPN Service Performance Monitoring [I-D.opsawg-yang-vpn-service-pm] defines a standard interface for performance management. In addition, distributed tracing mechanism defined in [W3C-Trace-Context] can also be used to follow, analyze and debug operations, such as configuration transactions, across multiple distributed systems.

However, alarm-centric solution described in [RFC8632] and performance-centric solution described in [I-D.opsawg-yang-vpn-service-pm], trace context-centric solution is based on a data source specific information and maintenance engineers' experience and fall short when keeping track of them separately in various different management systems, e.g., the frequency and quantity of alarms reported to Operating Support System (OSS) increased dramatically (in many cases multiple orders of magnitude) with the growth of service types and complexity, hard to aggregate in a single domain along with key performance metrics, various different events, notifications, overwhelm OSS platforms, result in low processing efficiency, inaccurate root cause identification and duplicated tickets.

Usually, the network modeling from device to different connection and service layers follows some existing standards. Once there are some failures happened on network devices, there could be some correlative alarms appeared on the upper layers. Theoretically, it is possible to compress a series of alarms into fewer incidents. The traditional working manner is also based on this correlation relationship. But the traditional working manner is time-consuming and labor-intensive which reduces efficiency. Additionally, it quite depends on the experience of maintenance engineers. Moreover, the investigation of some faults also depends on some other data like topology data or performance data. This complicates network troubleshooting, and the correlation of alarms and network services. Therefore, it is difficult to assess the impact of alarms on network services.

To address these challenges, an incident-centric solution is proposed, which also supports cross-domain or cross-layer root cause analysis and network troubleshooting. A network incident refers to an unexpected interruption of a network service, degradation of a network service quality, or sub-health of a network service while an alarm described in [RFC8632] represents an undesirable state in a resource that requires corrective actions. An alarm will always be reported when network resources are unexpected while an incident is reported only when network services are affected, e.g., symptoms (e.g.,CPU overloaded) at the device level defined in [I-D.opsawg-service-assurance-yang] or root cause alarms can be used to generate and report incidents when the network service is in sub-health state or gets degraded. An incident may be triggered by aggregation and analysis of multiple alarms or other network anomalies, for example, the protocols related to the interface fail to work properly due to the interface down, as a result, the network service becomes unavailable. An incident may also be raised through the analysis of some network performance metrics, for example, the delay or packet loss rate exceeds the threshold, causing degradation of the network service.

Artificial Intelligence (AI) and Machine Learning (ML) play a important role in the processing of large amounts of data with complex correlations. For example, Neural Network Algorithm or Hierarchy Aggregation Algorithm can be used to replace manual alarm correlation. Through online and offline learning, these algorithms can be continuously optimized to improve the efficiency of fault diagnosis.

This document defines the concepts, requirements, and architecture of incident management. The document also defines a YANG data model for incident lifecycle management, which improves troubleshooting efficiency, ensures network service quality, and improves network automation [RFC8969].

2. Terminology

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.

The following terms are defined in [RFC8632] are not redefined here:

The following terms are defined in this document:

Incident:

An unexpected interruption of a network service, degradation of network service quality, or sub-health of a network service.

Incident management:
Lifecycle management of incidents including incident identification, reporting, acknowledge, diagnosis, and resolution.
Incident management system:
An entity which implements incident management. It include incident management agent and incident management client.
Incident management agent:
An entity which provides some functions of incident management. For example, it can detect an incident, perform incident diagnosis, resolution and prediction,etc.
Incident management client:
An entity which can manage incidents. For example, it can receive incident notifications, query the information of incidents, instruct the incident management agent to diagnose, resolve, etc.

3. Sample Use Cases

3.1. Incident-Based Trouble Tickets dispatching

Currently, the dispatching of trouble tickets is mostly based on dispatching alarms. Some operators' maintenance engineers monitor and identify alarms which could link to the same fault. Then they dispatch these alarms to the same trouble ticket, which is in low automation. If there are many alarms, then the human costs are increased accordingly.

Some operators preset whitelist and adopt some coarse granularity association rules for the alarm management. It seems to improve fault management automation. However, some trouble tickets could be missed if the filtering conditions are too tight. If the filtering conditions are too loose, multiple trouble tickets would be dispatched to the same fault.

It is hard to achieve a perfect balance between the automation and duplicated trouble tickets under the traditional working situations. However, with the help of incident management, massive alarms can be aggregated into a few incidents, multiple trouble tickets will be saved. At the same time, incident management can keep high accuracy and automation. This could be an answer to this pain point of traditional trouble ticket dispatching

3.2. Fault Locating

Currently, to accomplish fault isolation and locating work, maintenance experts need to combine topology data, service data with huge amount of alarm data to do the analysis. Sometimes they also require some cooperation from the construction engineers who work on site, to operate fixing attempts on devices and then further investigation the root cause is required.

For example, for a common cable interruption, maintenance experts need to analyze the root cause alarm from massive alarms, and then trace the root alarm to the faulty span segment by segment. Next, site engineers perform tests at the source station to locate the interruption and locate the faulty optical exchange station. Then travel to the located optical exchange station to replace or splice fibers. During the whole process, multiple people are needed inside and outside the site.

With the help of incident management, the system can automatically locate the faulty span, and eliminate the need for manual analysis. By cooperating with the integrated OTDR within the equipment, we can determine the target optical exchange station before site visits. Multiple site visits and time are saved.

3.3. Fault Labelling

Fiber cutover is a common maintenance scenario for Operators. During the cutover process, maintenance experts must identify affected devices based on the cutover object and their experience. They will give these devices a mark to remind other maintenance engineers that it is not necessary to dispatch trouble tickets before the ending of cutover.

However, depending on human experience, it is very likely to make some mistakes. For example, some devices are missing to mark and some devices are marked incorrectly. If the devices are missing to mark, some trouble tickets will be dispatched during cutover, which are not needed actually. If the devices are wrongly marked, some fault not related to this cutover will be missing.

With incident management, maintenance experts only need to mark the cutover objects and do not need to mark the devices that would be affected. Because of the alarm aggregation capabilities and knowing the relationship between root cause alarm and correlative alarm, the fault management system can automatically identify correlative alarms, without dispatching any trouble tickets to the affected devices.

3.4. Energy Conservation

Under the global trend of energy conservation, emission reduction and safety management, more and more enterprises have joined the energy conservation and emission reduction ranks and adopted measures to turn off the power after work during non-working hours, making due contributions to the green earth. However, this proactive power-off measure periodically generates a large number of alarms on the network, and the traditional Operation and Management system can not effectively identify such non-real faults caused by the enterprise users? operations. Operators need to manually identify and rectify faults based on expert experience, wasting a large number of human resources.

Incident management can intelligently identify faults caused by periodic power-off on the tenant side and directly identify faults. As a result, operators do not need to dispatch trouble tickets for such faults any more, this can help to reduce human resource costs.

4. Incident Management Architecture


            +----------------------+-------------------+
            |                                          |
            |            Incident Management Client    |
            |                                          |
            |                                          |
            +------------+---------+---------+---------+
               ^         |         |         |
               |Incident |Incident |Incident |Incident
               |Report   |Ack      |Diagnose |Resolve
               |         |         |         |
               |         V         V         V
            +--+-------------------+---------+----------+
            |                                           |
            |                                           |
            |            Incident Management Agent      |
            |                                           |
            |                                           |
            |                                           |
            |                                           |
            +----------------------+-----+--+-----------+
                  ^       ^Abnormal         ^
                  |Alarm  |Operations       |Metrics
                  |Report |Report           |/Telemetry
                  |       |                 V
     +--------+-+-+-------+--------------++------------------+
     |                                                       |
     |                     Network                           |
     |                                                       |
     +------------------------------------+------------------+

Figure 1: Incident Management Architecture

Figure 1 illustrates the incident management architecture. Two key components for the incident management are incident management client and incident management agent.

Incident management agent can be deployed in network analytics platform, controllers or Orchestrators and provides functionalities such as incident detection, report, diagnosis, resolution, querying for incident lifecycle management.

Incident management client can be deployed in the network OSS or other business systems of operators and invokes the functionalities provided by incident management agent to meet the business requirements of fault management.

A typical workflow of incident management is as follows:

5. Functional Interface Requirements between the Client and the Agent

5.1. Incident Detection

In alarm-centric solution, although alarms are processed (based on manual rules or preconfigured rule) before being sent to the network OSS, multiple alarms are still sent to the network OSS. Whether these alarms have impact on network services and how much of the impact they created, it highly depends on the network OSS to analyze, which affects the efficiency of network maintenance.

        +--------------+
     +--|  Incident1   |
     |  +--+-----------+
     |     |  +-----------+
     |     +--+  alarm1   |
     |     |  +-----------+
     |     |
     |     |  +-----------+
     |     +--+  alarm2   |
     |     |  +-----------+
     |     |
     |     |  +-----------+
     |     +--+  alarm3   |
     |        +-----------+
     |  +--------------+
     +--|  Incident2   |
     |  +--+-----------+
     |     |  +-----------+
     |     +--+  metric1  |
     |     |  +-----------+
     |     |  +-----------+
     |     +--+  metric2  |
     |        +-----------+
     |
     |  +--------------+
     +--|  Incident3   |
        +--+-----------+
           |  +-----------+
           +--+ alarm1    |
           |  +-----------+
           |
           |  +-----------+
           +--| metric1   |
              +-----------+
Figure 2: Incident Detection

The incident management agent MUST be capable of detecting incidents. It can analyze the impact on network services from numerous alarms or monitor network service quality. Once the network service quality does not meet expectations, the incident agent MUST report the incident.

As described in Figure 2, multiple alarms, metrics, or hybrid can be aggregated into an incident after analysis. Each incident is associated with network services.

                     +----------------------+
                     |                      |
                     |     Orchestrator     |
                     |                      |
                     +----+-----------------+
                          ^VPN A Unavailable
                          |
                      +---+----------------+
                      |                    |
                      |     Controller     |
                      |                    |
                      |                    |
                      +-+-+-+-----+--+-----+
                      ^ ^            ^
                  IGP | |Interface   |IGP Peer
                 Down | |Down        | Abnormal
                      | |            |
     VPN A            | |            |
    +-----------------+-+------------+------------------*
    | \  +---+       ++-++         +-+-+        +---+  /|
    |  \ |   |       |   |         |   |        |   | / |
    |   \|PE1+-------| P1+X--------|P2 +--------|PE2|/  |
    |    +---+       +---+         +---+        +---+   |
    +---------------------------------------------------+

Figure 3: Example 1 of Incident Detection

As described in Figure 3, vpn a is deployed from PE1 to PE2, if a interface of P1 is going down, many alarms are triggered, such as interface down, igp down, and igp peer abnormal from P2. These alarms are aggregated and analyzed by controller, and the incident 'vpn unavailable' is triggered by the controller.

                    +----------------------+
                    |                      |
                    |     Orchestrator     |
                    |                      |
                    +----+-----------------+
                         ^VPN A Degradation
                         |
                     +---+----------------+
                     |                    |
                     |     controller     |
                     |                    |
                     |                    |
                     +-+-+-+-----+--+-----+
                       ^            ^
                       |Packet      |Path Delay
                       |Loss        |
                       |            |
    VPN A              |            |
   +-------------------+------------+-------------------+
   | \  +---+       ++-++         +-+-+        +---+  / |
   |  \ |   |       |   |         |   |        |   | /  |
   |   \|PE1+-------|P1 +---------|P2 +--------|PE2|/   |
   |    +---+       +---+         +---+        +---+    |
   +----------------------------------------------------+

Figure 4: Example 2 of Incident Detection

As described in Figure 4, controller collect the network metrics from network elements, it finds the packet loss of P1 and the path delay of P2 exceed the thresholds, an incident 'VPN A degradation' may be triggered after analysis.

5.2. Incident Diagnosis

After an incident is reported to the incident management client, the client MAY diagnose the incident to determine the root cause. Some diagnosis operations may affect the running network services. The client can choose not to perform that diagnosis operation after determining the impact is not trivial. The incident management agent can also perform self-diagnosis. However, the self-diagnosis MUST not affect the running network services. Possible diagnosis methods include link reachability detection, link quality detection, alarm/log analysis, and short-term fine-grained monitoring of network quality metrics, etc.

5.3. Incident Resolution

After the root cause is diagnosed, the client MAY resolve the incident. The client MAY choose resolve the incident by invoking other functions, such as routing calculation function, configuration function, dispatching a ticket or asking the agent to resolve it. Generally, the client would attempt to directly resolve the root cause. If the root cause cannot be resolved, an alternative solution SHOULD be required. For example, if an incident caused by a physical component failure, it cannot be automatically resolved, the standby link can be used to bypass the faulty component.

If the incident has been resolved, the client MAY indicate the agent to change the incident status to 'cleared'. If the incident is resolved by the agent, this indicator is unnecessary.

Incident resolution may affect the running network services. The client can choose not to perform those operations after determining the impact is not trivial.

6. Incident Data Model Concepts

6.1. Identifying the Incident Instance

An incident instance is associated with the specific network services instance and an incident name. An incident ID is used as an identifier of an incident instance, if an incident instance is detected, a new incident ID is created. The incident ID MUST be unique in the whole system.

6.2. The Incident Lifecycle

6.2.1. Incident Instance Lifecycle

From an incident instance perspective, an incident can have the following lifecycle: 'raised', 'updated', 'cleared'. When an incident is generated, the status is 'raised'. If the status changes after the incident is generated, (for example, self-diagnosis, diagnosis command issued by the client, or any other condition causes the status to change but does not reach the 'cleared' level.) , the status changes to 'updated'. When an incident is successfully resolved, the status changes to 'cleared'.

6.2.2. Operator Incident Lifecycle

From an operator perspective, the lifecycle of an incident instance includes 'acknowledged', 'diagnosed', and 'resolved'. When an incident instance is generated, the operator SHOULD acknowledge the incident. And then the operator attempts to diagnose the incident (for example, find out the root cause and affected components). Diagnosis is not mandatory. If the root cause and affected components are known when the incident is generated, diagnosis is not required. After locating the root cause and affected components, operator can try to resolve the incident.

7. Incident Data Model

7.1. Overview

   module: ietf-incident
     +--ro incidents
        +--ro incident* [incident-id]
           +--ro incident-id string
           +--ro csn uint64
           +--ro service-instance* string
           +--ro name string
           +--ro type enumeration
           +--ro domain identityref
           +--ro priority incident-priority
           +--ro status? enumeration
           +--ro ack-status? enumeration
           +--ro category identityref
           +--ro tenant? string
           +--ro detail? string
           +--ro resolve-suggestion? string
           +--ro sources
           | ...
           +--ro root-causes
           | ...
           +--ro events
           | ...
           +--ro raise-time? yang:date-and-time
           +--ro occur-time? yang:date-and-time
           +--ro clear-time? yang:date-and-time
           +--ro ack-time? yang:date-and-time
           +--ro last-updated? yang:date-and-time
     rpcs:
       +---x incident-acknowledge
       | ...
       +---x incident-diagnose
       | ...
       +---x incident-resolve
         ...
     notifications:
       +---n incident-notification
          +--ro incident-id? string
          ...

7.2. Incident Notifications

   notifications:
       +---n incident-notification
          +--ro incident-id? string
          +--ro csn uint64
          +--ro service-instance* string
          +--ro name string
          +--ro type enumeration
          +--ro domain identityref
          +--ro priority incident-priority
          +--ro status? enumeration
          +--ro ack-status? enumeration
          +--ro category identityref
          +--ro tenant? string
          +--ro detail? string
          +--ro resolve-suggestion? string
          +--ro sources
          |  +--ro source* [node]
          |     +--ro node
          |             -> /nw:networks/nw:network/nw:node/nw-inv:name
          |     +--ro resource* [name]
          |        +--ro name al:resource
          +--ro root-causes
          |  +--ro root-cause* [node]
          |     +--ro node
          |             -> /nw:networks/nw:network/nw:node/nw-inv:name
          |     +--ro resource* [name]
          |     |  +--ro name al:resource
          |     |  +--ro cause-name? string
          |     |  +--ro detail? string
          |     +--ro cause-name? string
          |     +--ro detail? string
          +--ro events
          |  +--ro event* [type original-node]
          |     +--ro type enumeration
          |     +--ro original-node union
          |     +--ro is-root? boolean
          |     +--ro (event-type-info)?
          |        +--:(alarm)
          |        |  +--ro alarm
          |        |     +--ro resource? leafref
          |        |     +--ro alarm-type-id? leafref
          |        |     +--ro alarm-type-qualifier? leafref
          |        +--:(notification)
          |        +--:(log)
          |        +--:(KPI)
          |        +--:(unknown)
          +--ro time? yang:date-and-time

A general notification, incident-notification, is provided here. When an incident instance is detected, the notification will be sent. After a notification is generated, if the incident management agent performs self diagnosis or the client uses the interfaces provided by the incident management agent to deliver diagnosis and resolution actions, the notification update behavior is triggered, for example, the root cause objects and affected objects are updated. When an incident is successfully resolved, the status of the incident would be set to 'cleared'.

7.3. Incident Acknowledge

    +---x incident-acknowledge
    |  +---w input
    |  |  +---w incident-id* string

After an incident is generated, updated, or cleared, (In some scenarios where automatic diagnosis and resolution are supported, the status of an incident may be updated multiple times or even automatically resolved.) The operator needs to confirm the incident to ensure that the client knows the incident.

The incident-acknowledge rpc can confirm multiple incidents at a time

7.4. Incident Diagnose

       +---x incident-diagnose
       |  +---w input
       |  |  +---w incident-id* string
       |  +--ro output
       |     +--ro incident* [incident-id]
       |        +--ro incident-id? string
       |        +--ro (result)?
       |           +--:(success)
       |           |  +--ro service-instance? string
       |           |  +--ro name? string
       |           |  +--ro domain? identityref
       |           |  +--ro priority? incident-priority
       |           |  +--ro impact? enumeration
       |           |  +--ro status? enumeration
       |           |  +--ro ack-status? enumeration
       |           |  +--ro category? identityref
       |           |  +--ro tenant? string
       |           |  +--ro detail? string
       |           |  +--ro resolve-suggestion? string
       |           |  +--ro sources
       |           |  |  +--ro source* [node]
       |           |  |     +--ro node? leafref
       |           |  |     +--ro resource* [name]
       |           |  |        +--ro name? al:resource
       |           |  +--ro root-causes
       |           |  |  +--ro root-cause* [node]
       |           |  |     +--ro node? leafref
       |           |  |     +--ro resource* [name]
       |           |  |     |  +--ro name? al:resource
       |           |  |     |  +--ro cause-name? string
       |           |  |     |  +--ro detail? string
       |           |  |     +--ro cause-name? string
       |           |  |     +--ro detail? string
       |           |  +--ro affects
       |           |  |  +--ro affect* [node]
       |           |  |     +--ro node? leafref
       |           |  |     +--ro resource* [name]
       |           |  |     |  +--ro name? al:resource
       |           |  |     |  +--ro state? enumeration
       |           |  |     |  +--ro detail? string
       |           |  |     +--ro state? enumeration
       |           |  |     +--ro detail? string
       |           |  +--ro links
       |           |  |  +--ro link* leafref
       |           |  +--ro events
       |           |  |  +--ro event* [type original-node]
       |           |  |     +--ro type? enumeration
       |           |  |     +--ro original-node? union
       |           |  |     +--ro is-root? boolean
       |           |  |     +--ro (event-type-info)?
       |           |  |        +--:(alarm)
       |           |  |        |  +--ro alarm
       |           |  |        |     +--ro resource? leafref
       |           |  |        |     +--ro alarm-type-id? leafref
       |           |  |        |     +--ro alarm-type-qualifier? leafref
       |           |  |        +--:(notification)
       |           |  |        +--:(log)
       |           |  |        +--:(KPI)
       |           |  |        +--:(unknown)
       |           |  +--ro time? yang:date-and-time
       |           +--:(failure)
       |              +--ro error-code? string
       |              +--ro error-message? string

After an incident is generated, incident diagnose rpc can be used to diagnose the incident and locate the root causes. Diagnosis can be performed on some detection tasks, such as BFD detection, flow detection, telemetry collection, short-term threshold alarm, configuration error check, or test packet injection.

If the diagnosis is successful, the latest status of the incident will be returned and a notification of the incident update will be triggered. If the diagnosis fails, error code and error message will be returned.

7.5. Incident Resolution

       +---x incident-resolve
          +---w input
          |  +---w incident* [incident-id]
          |     +---w incident-id
          |             -> /inc:incidents/inc:incident/inc:incident-id
          |     +---w resolved? empty
          +--ro output
             +--ro incident* [incident-id]
                +--ro incident-id string
                +--ro (result)?
                   +--:(success)
                   |  +--ro success? empty
                   |  +--ro time? yang:date-and-time
                   +--:(failure)
                      +--ro error-code? string
                      +--ro error-message? string

After the root cause and impact are determined, incident-resolve rpc can be used to resolve the incident (if the agent can resolve it) or indicate the incident instances have been resolved by other means. How to resolve an incident instance is out of the scope of this document.

Incident resolve rpc allows multiple incident instances to be resolved at a time. If an incident instance is successfully resolved, the success flag and resolve time will be returned, and a notification will be triggered to update the incident status to 'cleared'. If an incident fails to be resolved, an error code and an error message will be returned. If the incident content is changed during this process, a notification update will be triggered.

8. Incident Management YANG Module

<CODE BEGINS>
       file="ietf-incident@2023-03-13.yang"
   module ietf-incident {
     yang-version 1.1;
     namespace "urn:ietf:params:xml:ns:yang:ietf-incident";
     prefix inc;
     import ietf-yang-types {
       prefix yang;
       reference
         "RFC 6991: Common YANG Data Types";
     }
     import ietf-network {
       prefix nw;
       reference
         "RFC 8345: A YANG Data Model for Network Topologies";
     }
     import ietf-network-inventory {
       prefix nw-inv;
       reference
         "draft-wzwb-opsawg-network-inventory-management-01:
         An Inventory Management Model for Enterprise Networks";
     }
     import ietf-alarms {
       prefix al;
       reference
         "RFC 8632: A YANG Data Model for Alarm Management";
     }
     organization
       "IETF OPSAWG Working Group";
     contact
       "WG Web:   &lt;https://datatracker.ietf.org/wg/opsawg/&gt;
        WG List:  &lt;mailto:opsawg@ietf.org&gt;
        Author:   Chong Feng  &lt;mailto:frank.fengchong@huawei.com&gt;
        Author:   Tong Hu  &lt;mailto:hutong@cmhi.chinamobile.com&gt;
        Author:   Luis Miguel Contreras Murillo &lt;mailto:
                  luismiguel.contrerasmurillo@telefonica.com&gt;;
        Author :   Qin Wu   &lt;mailto:bill.wu@huawei.com&gt;
        Author:   ChaoDe Yu   &lt;mailto:yuchaode@huawei.com&gt;";

    description
       "This module defines the interfaces for incident management
        lifecycle.

        This module is intended for the following use cases:
        * incident lifecycle management:
          - incident report: report incident instance to client
                             when an incident instance is detected.
          - incident acknowledge: acknowledge an incident instance.
          - incident diagnose: diagnose an incident instance.
          - incident resolve: resolve an incident instance.

        Copyright (c) 2022 IETF Trust and the persons identified as
        authors of the code.  All rights reserved.

        Redistribution and use in source and binary forms, with or
        without modification, is permitted pursuant to, and subject
        to the license terms contained in, the Revised BSD License
        set forth in Section 4.c of the IETF Trust's Legal Provisions
        Relating to IETF Documents
        (https://trustee.ietf.org/license-info).
        This version of this YANG module is part of RFC XXXX; see the
        RFC itself for full legal notices.  ";
     revision 2023-03-13 {
       description "initial version";
       reference "RFC XXX: Yang module for incident management.";
     }
     //identities
     identity incident-domain {
       description "The abstract identity to indicate the domain of
                    an incident.";
     }
     identity single-domain {
       base incident-domain;
       description "single domain.";
     }
     identity access {
       base single-domain;
       description "access domain.";
     }
     identity ran {
       base access;
       description "radio access network domain.";
     }
     identity transport {
       base single-domain;
       description "transport domain.";
     }
     identity otn {
       base transport;
       description "optical transport network domain.";
     }
     identity ip {
       base single-domain;
       description "ip domain.";
     }
     identity ptn {
       base ip;
       description "packet transport network domain.";
     }

     identity cross-domain {
       base incident-domain;
       description "cross domain.";
     }

     identity incident-category {
       description "The abstract identity for incident category.";
     }
     identity device {
       base incident-category;
       description "device category.";
     }
     identity power-enviorment {
       base device;
       description "power system category.";
     }
     identity device-hardware {
       base device;
       description "hardware of device category.";
     }
     identity device-software {
       base device;
       description "software of device category";
     }
     identity line {
       base device-hardware;
       description "line card category.";
     }
     identity maintenance {
       base incident-category;
       description "maintenance category.";
     }
     identity network {
       base incident-category;
       description "network category.";
     }
     identity protocol {
       base incident-category;
       description "protocol category.";
     }
     identity overlay {
       base incident-category;
       description "overlay category";
     }
     identity vm {
       base incident-category;
       description "vm category.";
     }

     //typedefs
     typedef incident-priority {
       type enumeration {
         enum critical {
           description "the incident MUST be handled immediately.";
         }
         enum high {
           description "the incident should be handled as soon as
                        possible.";
         }
         enum medium {
           description "network services are not affected, or the
                        services are slightly affected,but corrective
                        measures need to be taken.";
         }
         enum low {
           description "potential or imminent service-affecting
                        incidents are detected,but services are
                        not affected currently.";
         }
       }
       description "define the priority of incident.";
     }
     typedef node-ref {
       type leafref {
         path "/nw:networks/nw:network/nw:node/nw-inv:name";
       }
       description "reference a network node.";
     }
     //groupings
     grouping resources-info {
       description "the grouping which defines the network
                    resources of a node.";
       leaf node {
         type node-ref;
         description "reference to a network node.";
       }
       list resource {
         key name;
         description "the resources of a network node.";
         leaf name {
            type al:resource;
            description "network resource name.";
         }
       }
     }

     grouping incident-time-info {
       description "the grouping defines incident time information.";
       leaf raise-time {
         type yang:date-and-time;
         description "the time when an incident instance is raised.";
       }
       leaf occur-time {
         type yang:date-and-time;
         description "the time when an incident instance is occured.
                      It's the occur time of the first event during
                      incident detection.";
       }
       leaf clear-time {
         type yang:date-and-time;
         description "the time when an incident instance is
                      resolved.";
       }
       leaf ack-time {
         type yang:date-and-time;
         description "the time when an incident instance is
                      acknowledged.";
       }
       leaf last-updated {
         type yang:date-and-time;
         description "the latest time when an incident instance is
                      updated";
       }
     }

     grouping incident-info {
       description "the grouping defines the information of an
                    incident.";
       leaf csn {
         type uint64;
         mandatory true;
         description "The sequence number of the incident instance.";
       }
       leaf-list service-instance {
         type string;
         description "the related network service instances of
                      the incident instance.";
       }
       leaf name {
         type string;
         mandatory true;
         description "the name of an incident.";
       }
       leaf type {
         type enumeration {
           enum fault {
             description "It indicates the type of the incident
                          is a fault, for example an interface
                          fails to work.";
           }
           enum potential-risk {
             description "It indicates the type of the incident
                          is a potential risk, for example high
                          CPU rate may cause a fault in the
                          future.";
           }
         }
         mandatory true;
         description "The type of an incident.";
       }
       leaf domain {
         type identityref {
           base incident-domain;
         }
         mandatory true;
         description "the domain of an incident.";
       }
       leaf priority {
         type incident-priority;
         mandatory true;
         description "the priority of an incident instance.";
       }

       leaf status {
         type enumeration {
           enum raised {
             description "an incident instance is raised.";
           }
           enum updated {
             description "the information of an incident instance
                          is updated.";
           }
           enum cleared {
             description "an incident is cleared.";
           }
         }
         default raised;
         description "The status of an incident instance.";
       }
       leaf ack-status {
         type enumeration {
           enum acknowledged;
           enum unacknowledged;
         }
         default unacknowledged;
         description "the acknowledge status of an incident.";
       }
       leaf category {
         type identityref {
           base incident-category;
         }
         mandatory true;
         description "The category of an incident.";
       }

       leaf tenant {
         type string;
         description "the identifier of related tenant.";
       }
       leaf detail {
         type string;
         description "detail information of this incident.";
       }
       leaf resolve-suggestion {
         type string;
         description "The suggestion to resolve this incident.";
       }
       container sources {
         description "The source components.";
         list source {
           key node;
           uses resources-info;
           min-elements 1;
           description "The source components of incident.";
         }
       }

       container root-causes{
         description "The root cause objects.";
         list root-cause {
           key node;
           description "the root causes of incident.";
           grouping root-cause-info {
             description "The information of root cause.";
             leaf cause-name {
               type string;
               description "the name of cause";
             }
             leaf detail {
               type string;
               description "the detail information of the cause.";
             }
           }
           uses resources-info {
             augment resource {
               description "augment root cause information.";
               //if root cause object is a resource of a node
               uses root-cause-info;
             }
           }
           //if root cause object is a node
           uses root-cause-info;
         }
       }
       container events {
         description "related event.";
         list event {
           key "type original-node";
           description "related event.";
           leaf type {
             type enumeration {
               enum alarm {
                 description "alarm type";
               }
               enum notification {
                 description "notification type";
               }
               enum log {
                 description "log type";
               }
               enum KPI {
                 description "KPI type";
               }
               enum unknown {
                 description "unknown type";
               }
             }
             description "event type.";
           }
           leaf original-node {
             type union {
               type node-ref;
               type empty;//self
             }
             description "the original node where the event occurs.";
           }
           leaf is-root {
             type boolean;
             default false;
             description "whether this event is the cause of
                           incident.";
           }
           choice event-type-info {
             description "event type information.";
             case alarm {
               when "type = 'alarm'";
               container alarm {
                 description "alarm type event.";
                 leaf resource {
                   type leafref {
                     path "/al:alarms/al:alarm-list/al:alarm"
                         +"/al:resource";
                   }
                   description "network resource.";
                   reference "RFC 8632: A YANG Data Model for Alarm
                              Management";
                 }
                 leaf alarm-type-id {
                   type leafref {
                     path "/al:alarms/al:alarm-list/al:alarm"
                         +"[al:resource = current()/../resource]"
                         +"/al:alarm-type-id";
                   }
                   description "alarm type id";
                   reference "RFC 8632: A YANG Data Model for Alarm
                               Management";
                 }
                 leaf alarm-type-qualifier {
                   type leafref {
                     path "/al:alarms/al:alarm-list/al:alarm"
                         +"[al:resource = current()/../resource]"
                         +"[al:alarm-type-id = current()/.."
                         +"/alarm-type-id]/al:alarm-type-qualifier";
                   }
                   description "alarm type qualitifier";
                   reference "RFC 8632: A YANG Data Model for Alarm
                              Management";
                 }
               }
             }
             case notification {
               //TODO
             }
             case log {
             //TODO
             }
             case KPI {
             //TODO
             }
             case unknown {
             //TODO
             }
           }
         }

       }

     }

     //data definitions
     container incidents {
       config false;
       description "the information of incidents.";
       list incident {
         key incident-id;
         description "the information of incident.";
         leaf incident-id {
           type string;
           description "the identifier of an incident instance.";
         }
         uses incident-info;
         uses incident-time-info;
       }
     }

     // notifications
     notification incident-notification {
       description "incident notification. It will be triggered when
                    the incident is raised, updated or cleared.";
       leaf incident-id {
         type string;
         description "the identifier of an incident instance.";
       }
       uses incident-info;
       leaf time {
         type yang:date-and-time;
         description "occur time of an incident instance.";
       }
     }
     // rpcs
     rpc incident-acknowledge {
       description "This rpc can be used to acknowledge the specified
                    incidents.";
       input {
         leaf-list incident-id {
           type string;
           description "the identifier of an incident instance.";
         }
       }
     }
     rpc incident-diagnose {
       description "This rpc can be used to diagnose the specified
                    incidents.";
       input {
         leaf-list incident-id {
           type string;
           description
             "the identifier of an incident instance.";
         }
       }
       output {
         list incident {
           key incident-id;
           description "The entry of returned incidents.";
           leaf incident-id {
             type string;
             description
               "the identifier of an incident instance.";
           }
           choice result {
             description "result information.";
             case success {
               uses incident-info;
               leaf time {
                 type yang:date-and-time;
                 description
                   "The update time of an incident.";
               }
             }
             case failure {
               leaf error-code {
                 type string;
                 description "error code";
               }
               leaf error-message {
                 type string;
                 description "error message";
               }
             }
           }
         }
       }
     }

     rpc incident-resolve {
       description "This rpc can be used to resolve the specified
                    incidents. It also can be used to set the
                    incident instances are resolved if these incident
                    instances are resolved by external system.";
       input {
         list incident {
           key incident-id;
           min-elements 1;
           description "incident instances.";
           leaf incident-id {
             type leafref {
               path "/inc:incidents/inc:incident/inc:incident-id";
             }
             description
               "the identifier of an incident instance.";
           }
           leaf resolved {
             type empty;
             description "indicate the incident instance has
                          been resolved.";
           }

         }
       }
       output {
         list incident {
           key incident-id;
           description "incident instances";
           leaf incident-id {
             type string;
             description "the identifier of incident instance";
           }
           choice result {
             description "result information";
             case success {
               leaf success {
                 type empty;
                 description "reslove incident instance
                              successfully";
               }
               leaf time {
                 type yang:date-and-time;
                 description "The resolved time of an incident.";
               }
             }
             case failure {
               leaf error-code {
                 type string;
                 description "error code";
               }
               leaf error-message {
                 type string;
                 description "error message.";
               }
             }
           }
         }
       }
     }
   }

<CODE ENDS>

9. IANA Considerations

9.1. The "IETF XML" Registry

This document registers one XML namespace URN in the 'IETF XML registry', following the format defined in [RFC3688].

URI: urn:ietf:params:xml:ns:yang:ietf-incident
Registrant Contact: The IESG.
XML: N/A, the requested URIs are XML namespaces.

9.2. The "YANG Module Names" Registry

This document registers one module name in the 'YANG Module Names' registry, defined in [RFC6020].

name: ietf-incident
prefix: inc
namespace: urn:ietf:params:xml:ns:yang:ietf-incident
RFC: XXXX
// RFC Ed.: replace XXXX and remove this comment

10. Security Considerations

The YANG modules specified in this document define a schema for data that is designed to be accessed via network management protocol such as NETCONF [RFC6241] or RESTCONF [RFC8040]. The lowest NETCONF layer is the secure transport layer, and the mandatory-to-implement secure transport is Secure Shell (SSH) [RFC6242]. The lowest RESTCONF layer is HTTPS, and the mandatory-to-implement secure transport is TLS [RFC8446].

The Network Configuration Access Control Model (NACM) [RFC8341] provides the means to restrict access for particular NETCONF or RESTCONF users to a preconfigured subset of all available NETCONF or RESTCONF protocol operations and content.

There are a number of data nodes defined in this YANG module that are writable/creatable/deletable (i.e., config true, which is the default). These data nodes may be considered sensitive or vulnerable in some network environments. Write operations (e.g., edit-config) to these data nodes without proper protection can have a negative effect on network operations. These are the subtrees and data nodes and their sensitivity/vulnerability:

Some of the readable data nodes in this YANG module may be considered sensitive or vulnerable in some network environments. It is thus important to control read access (e.g., via get, get-config, or notification) to these data nodes. These are the subtrees and data nodes and their sensitivity/vulnerability:

Some of the RPC operations in this YANG module may be considered sensitive or vulnerable in some network environments. It is thus important to control access to these operations. These are the operations and their sensitivity/vulnerability:

11. Contributors

Aihua Guo
Futurewei Technologies
Email: aihuaguo.ietf@gmail.com

12. Acknowledgments

The authors would like to thank Mohamed Boucadair, Zhidong Yin, Guoxiang Liu, Haomian Zheng, YuanYao for their valuable comments and great input to this work.

13. References

13.1. Normative References

[RFC2119]
Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, , <https://www.rfc-editor.org/info/rfc2119>.
[RFC3688]
Mealling, M., "The IETF XML Registry", BCP 81, RFC 3688, DOI 10.17487/RFC3688, , <https://www.rfc-editor.org/info/rfc3688>.
[RFC6020]
Bjorklund, M., Ed., "YANG - A Data Modeling Language for the Network Configuration Protocol (NETCONF)", RFC 6020, DOI 10.17487/RFC6020, , <https://www.rfc-editor.org/info/rfc6020>.
[RFC7950]
Bjorklund, M., Ed., "The YANG 1.1 Data Modeling Language", RFC 7950, DOI 10.17487/RFC7950, , <https://www.rfc-editor.org/info/rfc7950>.
[RFC8345]
Clemm, A., Medved, J., Varga, R., Bahadur, N., Ananthakrishnan, H., and X. Liu, "A YANG Data Model for Network Topologies", RFC 8345, DOI 10.17487/RFC8345, , <https://www.rfc-editor.org/info/rfc8345>.
[RFC8632]
Vallin, S. and M. Bjorklund, "A YANG Data Model for Alarm Management", RFC 8632, DOI 10.17487/RFC8632, , <https://www.rfc-editor.org/info/rfc8632>.

13.2. Informative References

[I-D.ietf-opsawg-yang-vpn-service-pm]
Wu, B., Wu, Q., Boucadair, M., de Dios, O. G., and B. Wen, "A YANG Model for Network and VPN Service Performance Monitoring", Work in Progress, Internet-Draft, draft-ietf-opsawg-yang-vpn-service-pm-15, , <https://datatracker.ietf.org/doc/html/draft-ietf-opsawg-yang-vpn-service-pm-15>.
[I-D.wzwb-opsawg-network-inventory-management]
Wu, B., Zhou, C., Wu, Q., and M. Boucadair, "An Inventory Management Model for Enterprise Networks", Work in Progress, Internet-Draft, draft-wzwb-opsawg-network-inventory-management-01, , <https://datatracker.ietf.org/doc/html/draft-wzwb-opsawg-network-inventory-management-01>.
[RFC8174]
Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, , <https://www.rfc-editor.org/info/rfc8174>.
[RFC8969]
Wu, Q., Ed., Boucadair, M., Ed., Lopez, D., Xie, C., and L. Geng, "A Framework for Automating Service and Network Management with YANG", RFC 8969, DOI 10.17487/RFC8969, , <https://www.rfc-editor.org/info/rfc8969>.
[W3C-Trace-Context]
W3C, "W3C Recommendation on Trace Context", , <https://www.w3.org/TR/2021/REC-trace-context-1-20211123/>.

Authors' Addresses

Chong Feng (editor)
Huawei
101 Software Avenue, Yuhua District
Nanjing
Jiangsu, 210012
China
Tong Hu
China Mobile (Hangzhou) Information Technology Co., Ltd
Building A01, 1600 Yuhangtang Road, Wuchang Street, Yuhang District
Hangzhou
ZheJiang, 311121
China
Luis Miguel Contreras Murillo
Telefonica I+D
Madrid
Spain
Qin Wu
Huawei
101 Software Avenue, Yuhua District
Nanjing
Jiangsu, 210012
China
Chaode Yu
Huawei