How are NSP assurance events retrieved and recorded?

Assurance events

Assurance events are used as a troubleshooting tool to give the user insight into the events that led to a certain state. For example, an operator could examine events that occurred on an NE or service prior to a critical alarm and see if there was a configuration change that resulted in an alarm.

Assurance events are sequential historical events recorded on NE or service objects and their hierarchical sub-components. The type of events that can generate an assurance event are:

Event recording is handled by NFM-P for NFM-P-managed objects, while NSP-managed object recording is handled in NSPOS by the timedevent-recording-app pod. An NSP administrator enables event logging globally for NSP and NFM-P in the Network Assurance Settings; see How do I manage assurance event recording? Any changes to the NFM-P event recording policy made directly in the NFM-P GUI will have no effect on the NSP event recording settings.

Assurance event recording in NSPOS

For objects managed by MDM or WS-NOC, event logging is implemented using time series model recording. .

NSPOS records events for the following objects:

Network Element

Shelf

Slot

Card

Port

Lag

Physical Link

Physical Link Site

Physical Link Endpoint

Service

Service Site

Service Endpoint

Service Tunnel Binding

Tunnel

LSP

LSP Path

Assurance event format

The following is a sample NSPOS event:

{
  "nsp-model-notification:object-creation": {
    "schema-nodeid": "/nsp-equipment:network/network-element/hardware-component/container",
    "instance-id": "/nsp-equipment:network/network-element[ne-id='xx.xx.xx.xx']/hardware-component/container[component-id='shelf=1/sfmSlot=2']",
    "context": "nsp-db-synchronizer",
    "tree": {
      "/nsp-equipment:network/network-element/hardware-component/container": {
        "@": {
          "nsp-model:schema-nodeid": "/nsp-equipment:network/network-element/hardware-component/container",
          "nsp-model:identifier": "/nsp-equipment:network/network-element[ne-id='xx.xx.xx.xx']/hardware-component/
     container[component-id='shelf=1/sfmSlot=2']",
          "nsp-model:creation-time": "20xx-04-17T13:08:05.968Z",
          "nsp-model:last-modified-time": "20xx-04-17T13:08:05.968Z",
          "nsp-model:sources": [
            "fdn:app:mdm-ami-cmodel:16.16.16.16:equipment:EquipmentHolder:/sfm[sfm-slot='2']"
          ]
        },
        "mac-address": null,
        "mfg-deviations": null,
        "mfg-name": null,
        "availability-state": [],
        "oper-state": "disabled",
        "description": null,
        "mfg-date": "",
        "mfg-assembly-number": "",
        "hw-mac-address": null,
        "admin-state": "unlocked",
        "actual-type": "unassigned",
        "parent-rel-pos": 2,
        "ne-id": "xx.xx.xx.xx"
    "serial-num": "",
        "clei-code": null,
        "holder-state": "not-expected-not-installed",
        "component-id": "shelf=1/sfmSlot=2",
        "ne-name": "Assurance-MDM-2",
        "provisioned-type": "unassigned",
        "part-number": "",
        "standby-state": "providing-service",
        "name": "SFM Slot-2(unassigned)",
        "state-reason": [],
        "class": "container"
      }
    },
    "event-time": "20xx-17T13:08:06.695604Z"
  }
}

The following is a sample NFM-P event:

{
  "eventId": "294a9268-b926-40d9-a08f-c71d6c73a0d2",
  "samUser": "admin",
  "timeIssued": 1713384849182,
  "objectDisplayName": "Card - 1",
  "objectClass": "equipment.BaseCard",
  "changes": [
    {
      "newValue": "outOfService",
      "propertyName": "administrativeState",
      "oldValue": "inService"
    }
  ],
  "objectName": "network:xx.xx.xx.xx:shelf-1:cardSlot-1:card",
  "eventType": "ConfigurationEvent",
  "version": "1.0"
}

Each NFM-P or NSPOS event contains the following information:

Assurance event recording in NFM-P

Assurance events for NFM-P-managed objects are recorded using NFM-P event generation (JMS). The assurance event recorder subscribes to the JMS events. If an event is on an object of interest, it is translated into an assurance event record and logged using the event logging framework.

By default, assurance events are recorded in an Oracle database, but if the customer has configured an auxiliary database, the recording is automatically routed to the auxiliary database.

An NFM-P Event policy allows you to enable/disable event recording (Admin State Up/Down) and the log retention time. To configure an Event policy, open NFM-P, click Tools menu, Events, Event Policies. The framework also provides tools to purge events.

Event log retention time defaults and minimum/maximum values depend on the type of database you are using. 

Table 2-3: Event log retention time defaults and minimum/maximum values

Database type

Default retention time

Minimum retention time

Maximum retention time

Oracle

168 hours (one week)

One hour

720 hours (1one month)

Auxiliary

720 hours (one month)

One day

8760 hours (one year)

In addition to NFM-P Event Policy configuration, you must enable event recording in the Timeline Settings form. Click png5.png More, Timeline Settings. You can enable event recording for specific objects:

Table 2-4: Objects with event recording

Object type

NFM-P class

Network Element

netw.NetworkElement

Link

netw.AbstractPhysicalLink

Card

equipment.Card

Port

equipment.Port

Site

rtr.ProtocolSite

Network Interface

rtr.NetworkInterface

LAG Interface

lag.Interface

Site Sync

sonet.SiteSync

MPLS

rsvp.Interface

IGP

isis.Interface

IGP

ospf.Interface

VNF Instance

nfv.VNFInstance

VNF Component

nfv.VNFComponent

Assurance events retrieval

NSP retrieves events from the appropriate database (NFM-P or NSPOS), based on the original source of the object.

Assurance events are retrieved using API commands from the web component library (assurance-share-md), which provides web components to display events on a timeline. Assurance events are recorded and stored on a server that is accessible from NSP using the web component library.

The web component library:

The Event Timeline displays events for multiple objects at the same time if all of the selected objects are part of the same object hierarchy, and all of the objects are at the same level in the hierarchy (for example, multiple service sites for the same service). As the number of related/child objects to a troubleshooting object increases, response times for retrieved events will also increase.

Assurance event purging

An NSP administrator can purge event records in the Network Assurance Settings; see How do I manage assurance event recording? Event purging only removes events stored in the NSP PostgreSQL database; not events recorded and stored in NFM-P. NFM-P event purging is performed from the assurance event recording policy in the NFM-P GUI.