Internet DRAFT - draft-feng-opsawg-incident-management
draft-feng-opsawg-incident-management
OPSAWG C. Feng, Ed.
Internet-Draft Huawei
Intended status: Standards Track T. Hu
Expires: 14 September 2023 CMCC
LM. Contreras
Telefonica I+D
Q. Wu
C. Yu
Huawei
13 March 2023
Incident Management for Network Services
draft-feng-opsawg-incident-management-00
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.
Copyright Notice
Copyright (c) 2023 IETF Trust and the persons identified as the
document authors. All rights reserved.
Feng, et al. Expires 14 September 2023 [Page 1]
Internet-Draft Incident Management March 2023
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents (https://trustee.ietf.org/
license-info) in effect on the date of publication of this document.
Please review these documents carefully, as they describe your rights
and restrictions with respect to this document. Code Components
extracted from this document must include Revised BSD License text as
described in Section 4.e of the Trust Legal Provisions and are
provided without warranty as described in the Revised BSD License.
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 4
3. Sample Use Cases . . . . . . . . . . . . . . . . . . . . . . 5
3.1. Incident-Based Trouble Tickets dispatching . . . . . . . 5
3.2. Fault Locating . . . . . . . . . . . . . . . . . . . . . 6
3.3. Fault Labelling . . . . . . . . . . . . . . . . . . . . . 6
3.4. Energy Conservation . . . . . . . . . . . . . . . . . . . 7
4. Incident Management Architecture . . . . . . . . . . . . . . 7
5. Functional Interface Requirements between the Client and the
Agent . . . . . . . . . . . . . . . . . . . . . . . . . . 9
5.1. Incident Detection . . . . . . . . . . . . . . . . . . . 9
5.2. Incident Diagnosis . . . . . . . . . . . . . . . . . . . 12
5.3. Incident Resolution . . . . . . . . . . . . . . . . . . . 13
6. Incident Data Model Concepts . . . . . . . . . . . . . . . . 13
6.1. Identifying the Incident Instance . . . . . . . . . . . . 13
6.2. The Incident Lifecycle . . . . . . . . . . . . . . . . . 13
6.2.1. Incident Instance Lifecycle . . . . . . . . . . . . . 13
6.2.2. Operator Incident Lifecycle . . . . . . . . . . . . . 14
7. Incident Data Model . . . . . . . . . . . . . . . . . . . . . 14
7.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . 14
7.2. Incident Notifications . . . . . . . . . . . . . . . . . 15
7.3. Incident Acknowledge . . . . . . . . . . . . . . . . . . 17
7.4. Incident Diagnose . . . . . . . . . . . . . . . . . . . . 17
7.5. Incident Resolution . . . . . . . . . . . . . . . . . . . 19
8. Incident Management YANG Module . . . . . . . . . . . . . . . 19
9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 32
9.1. The "IETF XML" Registry . . . . . . . . . . . . . . . . . 32
9.2. The "YANG Module Names" Registry . . . . . . . . . . . . 32
10. Security Considerations . . . . . . . . . . . . . . . . . . . 32
11. Contributors . . . . . . . . . . . . . . . . . . . . . . . . 33
12. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 33
13. References . . . . . . . . . . . . . . . . . . . . . . . . . 33
13.1. Normative References . . . . . . . . . . . . . . . . . . 33
13.2. Informative References . . . . . . . . . . . . . . . . . 34
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 34
Feng, et al. Expires 14 September 2023 [Page 2]
Internet-Draft Incident Management March 2023
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
Feng, et al. Expires 14 September 2023 [Page 3]
Internet-Draft Incident Management March 2023
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:
* alarm
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.
Feng, et al. Expires 14 September 2023 [Page 4]
Internet-Draft Incident Management March 2023
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
Feng, et al. Expires 14 September 2023 [Page 5]
Internet-Draft Incident Management March 2023
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.
Feng, et al. Expires 14 September 2023 [Page 6]
Internet-Draft Incident Management March 2023
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
Feng, et al. Expires 14 September 2023 [Page 7]
Internet-Draft Incident Management March 2023
+----------------------+-------------------+
| |
| 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:
Feng, et al. Expires 14 September 2023 [Page 8]
Internet-Draft Incident Management March 2023
* Some alarms or abnormal operations, network performance metrics
are reported from the network. Incident management agent receives
these alarms/abnormal operations/metrics and analyzes the impact
of these alarms on network services. If the analysis result
indicates that network services are affected, an incident will be
reported to the client.
* Incident management client receives the incident raised by agent,
and acknowledge it. Client may invoke the 'incident diagnose' rpc
to diagnose this incident to find the root causes.
* If the root causes have been found, the client can resolve this
incident by invoking the 'incident resolve' rpc operation,
dispatching a ticket or using other functions (e.g. routing
calculation,configuration)
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.
Feng, et al. Expires 14 September 2023 [Page 9]
Internet-Draft Incident Management March 2023
+--------------+
+--| 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.
Feng, et al. Expires 14 September 2023 [Page 10]
Internet-Draft Incident Management March 2023
+----------------------+
| |
| 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.
Feng, et al. Expires 14 September 2023 [Page 11]
Internet-Draft Incident Management March 2023
+----------------------+
| |
| 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.
Feng, et al. Expires 14 September 2023 [Page 12]
Internet-Draft Incident Management March 2023
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'.
Feng, et al. Expires 14 September 2023 [Page 13]
Internet-Draft Incident Management March 2023
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
Feng, et al. Expires 14 September 2023 [Page 14]
Internet-Draft Incident Management March 2023
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
Feng, et al. Expires 14 September 2023 [Page 15]
Internet-Draft Incident Management March 2023
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
Feng, et al. Expires 14 September 2023 [Page 16]
Internet-Draft Incident Management March 2023
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
Feng, et al. Expires 14 September 2023 [Page 17]
Internet-Draft Incident Management March 2023
| | | +--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.
Feng, et al. Expires 14 September 2023 [Page 18]
Internet-Draft Incident Management March 2023
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;
Feng, et al. Expires 14 September 2023 [Page 19]
Internet-Draft Incident Management March 2023
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: <https://datatracker.ietf.org/wg/opsawg/>
WG List: <mailto:opsawg@ietf.org>
Author: Chong Feng <mailto:frank.fengchong@huawei.com>
Author: Tong Hu <mailto:hutong@cmhi.chinamobile.com>
Author: Luis Miguel Contreras Murillo <mailto:
luismiguel.contrerasmurillo@telefonica.com>;
Author : Qin Wu <mailto:bill.wu@huawei.com>
Author: ChaoDe Yu <mailto:yuchaode@huawei.com>";
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
Feng, et al. Expires 14 September 2023 [Page 20]
Internet-Draft Incident Management March 2023
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.";
}
Feng, et al. Expires 14 September 2023 [Page 21]
Internet-Draft Incident Management March 2023
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 {
Feng, et al. Expires 14 September 2023 [Page 22]
Internet-Draft Incident Management March 2023
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.";
Feng, et al. Expires 14 September 2023 [Page 23]
Internet-Draft Incident Management March 2023
}
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.";
Feng, et al. Expires 14 September 2023 [Page 24]
Internet-Draft Incident Management March 2023
}
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.";
}
Feng, et al. Expires 14 September 2023 [Page 25]
Internet-Draft Incident Management March 2023
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 {
Feng, et al. Expires 14 September 2023 [Page 26]
Internet-Draft Incident Management March 2023
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 {
Feng, et al. Expires 14 September 2023 [Page 27]
Internet-Draft Incident Management March 2023
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
Feng, et al. Expires 14 September 2023 [Page 28]
Internet-Draft Incident Management March 2023
}
}
}
}
}
//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.";
}
}
Feng, et al. Expires 14 September 2023 [Page 29]
Internet-Draft Incident Management March 2023
}
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
Feng, et al. Expires 14 September 2023 [Page 30]
Internet-Draft Incident Management March 2023
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";
}
Feng, et al. Expires 14 September 2023 [Page 31]
Internet-Draft Incident Management March 2023
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.
Feng, et al. Expires 14 September 2023 [Page 32]
Internet-Draft Incident Management March 2023
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, March 1997,
<https://www.rfc-editor.org/info/rfc2119>.
[RFC3688] Mealling, M., "The IETF XML Registry", BCP 81, RFC 3688,
DOI 10.17487/RFC3688, January 2004,
<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, October 2010,
<https://www.rfc-editor.org/info/rfc6020>.
Feng, et al. Expires 14 September 2023 [Page 33]
Internet-Draft Incident Management March 2023
[RFC7950] Bjorklund, M., Ed., "The YANG 1.1 Data Modeling Language",
RFC 7950, DOI 10.17487/RFC7950, August 2016,
<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, March
2018, <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, September
2019, <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, 11 November 2022,
<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, 10 February 2023,
<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,
May 2017, <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,
January 2021, <https://www.rfc-editor.org/info/rfc8969>.
[W3C-Trace-Context]
W3C, "W3C Recommendation on Trace Context", 23 November
2021, <https://www.w3.org/TR/2021/REC-trace-context-
1-20211123/>.
Authors' Addresses
Feng, et al. Expires 14 September 2023 [Page 34]
Internet-Draft Incident Management March 2023
Chong Feng (editor)
Huawei
101 Software Avenue, Yuhua District
Nanjing
Jiangsu, 210012
China
Email: frank.fengchong@huawei.com
Tong Hu
China Mobile (Hangzhou) Information Technology Co., Ltd
Building A01, 1600 Yuhangtang Road, Wuchang Street, Yuhang District
Hangzhou
ZheJiang, 311121
China
Email: hutong@cmhi.chinamobile.com
Luis Miguel Contreras Murillo
Telefonica I+D
Madrid
Spain
Email: luismiguel.contrerasmurillo@telefonica.com
Qin Wu
Huawei
101 Software Avenue, Yuhua District
Nanjing
Jiangsu, 210012
China
Email: bill.wu@huawei.com
Chaode Yu
Huawei
Email: yuchaode@huawei.com
Feng, et al. Expires 14 September 2023 [Page 35]
