IBM Support

ServiceNow Integration scenarios and Best Practices with IBM Cloud Pak for AIOps

Question & Answer


Question

Servicenow integration with AIOps scenarios :

a) Is the ServiceNow connector limited to ticketing only for incidents, or can it also be used for ticketing alerts?
b) Does the ServiceNow connector support XML communication? The documentation references REST API and JSONATA Mapping. If ServiceNow integration utilizes SOAP/XML endpoints or payloads, can it integrate with AIOps?
c) What is the purpose of the ServiceNow App? Is it intended for ServiceNow to connect with IBM Cloud Pak for AIOps? How does it differ from the connector?
d) When triggering ticket creation, does the JSONATA mapping in the connector define the payload sent to ServiceNow? Can you provide a detailed example payload for reference?
e) How are ticket updates in ServiceNow tracked back to AIOps? Does the connector automatically resync updates, or is additional configuration (such as Impact policies or custom coding) required?
f) Are the workflows for tickets and incidents independent or dependent? Does the incident lifecycle (e.g., in progress, closed, assigned, resolved) impact or associate with the ticket lifecycle?
g) Is there an auto-resync facility available between incident notes and the worklog of ServiceNow tickets when a new alert is added, an alert is cleared, or when an incident is closed/resolved?
h) What is the maximum tested parallel request capacity handled by the connector in a ServiceNow integration?
i) Can the same connector be used for ticket analysis and change risk assessment? Is it necessary to maintain identical payloads for ticket creation and similar incidents, or should multiple instances of the connector be deployed for each purpose?
j) Can the connector support bi-directional integration, allowing updates from both AIOps and ServiceNow to be synchronized automatically?

k)  For the sample payload provided for JSONATA Mapping, is the provided payload the default mapping defined within the connector for every candidate incident configured for ticketing?

l) If you would like to configure Impact, Urgency, and Assignee/Assignment Group during ticket creation. For example, if there is an incident generated for a Link Down alert, and you want the ticket to be created with Impact = High, Urgency = 2-High, and Assignee Group = "IPCore". How can they configure this in the mapping?

 

m) Since we can use same connector for ticket analysis and change risk assessment, so can the same connector be used for ticket creation (for AIOps incidents) or do we need to have a separate connector exclusively for ticket creations/ updates and another one for ticket analysis and change risk?

n) Similar to netcool where the alarm is cleared but the service now ticket is not closed, alarm remains in the system until we receive the close status updated back to alarm. How can we control the same in AIOps?

o) In case if we need to reopen the service now ticket, how can we pass the updates (severity change, aiops incident status or any other parameters) to service now and will service now consider all these changes while reopening the ticket

 

p) While two way sync is enabled, if the update is not synchronized back to AIOps, will it try again during next sync or will the failed updates be ignored?

Cause

Product Version: AIOps 4.12.0

The customer was assisted in understanding the ServiceNow integration with AIOps through detailed explanations and guidance on various aspects of the integration, including:

  1. Confirming the ServiceNow application version (4.4.2) compatibility with IBM Cloud Pak for AIOps version 4.12.0.
  2. Clarifying that the ServiceNow connector can be used for both incident and alert ticketing.
  3. Explaining the use of JSONata expressions for mapping AIOps fields to ServiceNow incident fields.
  4. Detailing the synchronization of updates between AIOps and ServiceNow, including enabling two-way sync by setting specific properties in the integration configuration.
  5. Addressing how to configure additional fields like Impact, Urgency, and Assignee during ticket creation using incident mapping and AIOps policies.
  6. Confirming the same connector can be used for ticket analysis, change risk assessment, and ticket creation/updates.
  7. Explaining the handling of updates and re-sync mechanisms between AIOps and ServiceNow.

Answer

a) Is the ServiceNow connector limited to ticketing only for incidents, or can it also be used for ticketing alerts?

The ServiceNow connector can also be used for ticketing alerts. Alerts are mapped to incidents. For information on mapping alert data to ServiceNow incidents, refer to the documentation: IBMCloud Pak for AIOps ServiceNow integration documentation

Alert records in ServiceNow are created only when they are associated with an incident. An AIOps alert that does not have a corresponding AIOps incident cannot be ticketed in ServiceNow.

b) Does the ServiceNow connector support XML communication? The documentation references REST API and JSONATA mapping. If the ServiceNow integration utilizes SOAP/XML endpoints or payloads, can it integrate with AIOps?

No, the ServiceNow connector does not support XML communication.

c) What is the purpose of the ServiceNow App? Is it intended for ServiceNow to connect with IBM Cloud Pak for AIOps? How does it differ from the connector?

For the Cloud Pak for AIOps ServiceNow integration to function properly, the ServiceNow instance must have a dedicated app/plugin installed from the ServiceNow Store. This plugin enables all the necessary features required for seamless communication between ServiceNow and Cloud Pak for AIOps (CP4AIOps), such as the Change Risk Assessment tab and related capabilities.

In essence, the plugin acts as the communication and translation layer between the ServiceNow instance and the CP4AIOps ServiceNow integration.

d) When triggering ticket creation, does the JSONATA mapping in the connector define the payload sent to ServiceNow? Can you provide a detailed example payload for reference?

No, the JSONATA mapping does not define the entire payload sent to ServiceNow. Cloud Pak for AIOps sends a complete payload containing incident data, and the JSONATA mapping is only an optional object within that payload. It is used to add custom field values to the ServiceNow incident after the base fields are set.

The complete payload sent to ServiceNow includes incident and alert data and is not customizable.

Sample Payload:

{ "specversion": "1.0", "type": "com.ibm.sdlc.snow.incident.create", "connectionid": "e0d8767c-e807-4f71-bdb5-4d5144831c68", "componentname": "connector", "data": { "story_id": "7677d506-7d3b-420d-a4e6-8009d063216d", "snow_payload": { "story": { "title": "testDebugsnowDebug1", "id": "7677d506-7d3b-420d-a4e6-8009d063216d", "number": "dnfv-bn07", "description": "testDebugsnowDebug1", "priority": 5, "detail": { "topologyDefinedApplications": [] }, "associatedAlerts": [ { "id": "5d0c64c0-26bc-4130-aa7d-99630a9a2666", "type": { "classification": "testDebugsnowDebug1 critical server impact", "eventType": "problem" }, "severity": "2", "state": "open", "first_occurrence_timestamp": "2024-08-07T18:06:21", "last_occurrence_timestamp": "2024-08-07T18:06:21", "lifecycleState": { "lastStateChangeTime": "2024-08-07T18:06:30", "value": "open" }, "summary": "testDebugsnowDebug1", "name": "testDebugsnowDebug1", "resource": { "name": "testDebugsnowDebug1 server 1", "type": "testDebugsnowDebug1 server", "cluster": "democluster", "interface": "devtest", "controller": "devscntrl", "connectionId": "f5aa7fa9-92eb-4bec-942c-37eb3e3e9601", "location": "30.2672° N, 97.7431° W", "service": "testDebugsnowDebug1 devtralf", "hostname": "devs.laptop.for.testing", "ipAddress": "1.9.99.999" }, "event_count": "1", "details": { "current_state": "open", "condition_id": "7973772", "condition_name": "Replication currency Within 8 Hours above threshold", "event_type": "INCIDENT", "incident_id": "4134482", "version": "1.0", "cluster": "ipctest", "owner": "Marketing", "clusterName": "Big Cluster", "cloud": "IBM", "vendor": "ICP", "environment": "Prod", "region": "US", "datacenter": "toronto", "location": "freeland" }, "deduplicationKey": "{cluster=democluster, connectionId=f5aa7fa9-92eb-4bec-942c-37eb3e3e9601, controller=devscntrl, hostname=devs.laptop.for.testing, interface=devtest, ipAddress=1.9.99.999, location=30.2672° N, 97.7431° W, name=testDebugsnowDebug1 server 1, service=testDebugsnowDebug1 devtralf, type=testDebugsnowDebug1 server}-testDebugsnowDebug1 critical server impact-", "targetUrl": "https://cpd-katamari.apps.bvt-svl-05122014.cp.fyre.ibm.com/aiops/cfd95b7e-3bc7-4006-a4a8-a73a79c71255/resolution-hub/incidents/all/7677d506-7d3b-420d-a4e6-8009d063216d/alerts", "snowSysId": "" } ], "topologyLink": "", "storyLink": "https://cpd-katamari.apps.bvt-svl-05122014.cp.fyre.ibm.com/aiops/cfd95b7e-3bc7-4006-a4a8-a73a79c71255/resolution-hub/incidents/all/7677d506-7d3b-420d-a4e6-8009d063216d/overview", "state": "OPEN" } } }, "source": "/chatops-orchestrator", "id": "47ea1628-0082-41ca-a6e3-fa7dafd94f98" }e) How are ticket updates in ServiceNow tracked back to AIOps? Does the connector automatically resync updates, or is additional configuration (such as Impact policies or custom coding) required?

By default, updates made in AIOps are synchronized to ServiceNow (one-way sync), and this behaviour is not configurable.

To enable synchronization of updates from ServiceNow back to AIOps (two-way sync), the toggle option "Update incidents in Cloud Pak for AIOps when corresponding incidents are updated in ServiceNow" must be set to ON while creating the integration. Once enabled, the connector continuously polls ServiceNow at regular intervals and automatically resynchronizes incident updates.

f) Are the workflows for tickets and incidents independent or dependent? Does the incident lifecycle (e.g., in progress, closed, assigned, resolved) impact or associate with the ticket lifecycle?

In ServiceNow, tickets and incidents are the same entity, so there is no separate workflow. They share the same lifecycle.

Yes, the incident lifecycle directly impacts the integration workflow. By default, updates made in AIOps are synced to ServiceNow (one-way sync), and this behavior is not configurable.

When "Update incidents in Cloud Pak for AIOps when corresponding incidents are updated in ServiceNow" is set to ON, all lifecycle state changes (such as in progress, assigned, resolved, and closed) are continuously synchronized back to AIOps for real-time monitoring (two-way sync).

g) Is there an auto-resync facility available between incident notes and the worklog of ServiceNow tickets when a new alert is added, an alert is cleared, or when an incident is closed/resolved?

The integration syncs incident data and alerts, but it does not provide automatic resynchronization of work notes or comments when alerts change or incidents are closed.

h) What is the maximum tested parallel request capacity handled by the connector in a ServiceNow integration?

The connector has been tested to handle up to 100 parallel alert requests per second, including create, update, and delete operations, in a ServiceNow integration.

i) Can the same connector be used for ticket analysis and change risk assessment? Is it necessary to maintain identical payloads for ticket creation and similar incidents, or should multiple instances of the connector be deployed for each purpose?

Yes, a single connector instance can handle both purposes simultaneously. The connector runs multiple polling workflows concurrently, pulling incidents for ticket analysis/ML training and change requests for change risk assessment without requiring separate deployments.

j) Can the connector support bi-directional integration, allowing updates from both AIOps and ServiceNow to be synchronized automatically?

Yes, the connector supports bi-directional integration. By default, updates made in AIOps are synced to ServiceNow (one-way sync), and this behavior is not configurable.

To enable bi-directional integration, set "Update incidents in Cloud Pak for AIOps when corresponding incidents are updated in ServiceNow" to ON, so that changes made in ServiceNow are synchronized back to AIOps (two-way sync).


 k)  For the sample payload provided for JSONATA Mapping, is the provided payload the default mapping defined within the connector for every candidate incident configured for ticketing?

No, the sample payload shared is not the default mapping. The default mapping is built into the connector and is not configurable. The incident mapping functionality that we provide is an additional layer on top of the default mapping. The payload shared is only intended as a sample example.

l) If you would like to configure Impact, Urgency, and Assignee/Assignment Group during ticket creation. For example, if there is an incident generated for a Link Down alert, they want the ticket to be created with Impact = High, Urgency = 2-High, and Assignee Group = "IPCore". How can you configure this in the mapping?


Impact and Urgency:

Mapping these fields through the incident mapping can be challenging because they are managed by AIOps and may be overwritten during processing. Instead, it is recommended to configure the ticket priority through the AIOps policy. When creating or updating an AIOps policy, the Priority field can be used to set the desired ticket priority.

Assignee and Assignment Group:

Assignee and Assignment Group can be configured in the Incident Field Mapping while creating the ServiceNow integration by specifying the corresponding ServiceNow sys_ids of the user and/or group to which the incident should be assigned.

 

m) Since we can use same connector for ticket analysis and change risk assessment, so can the same connector be used for ticket creation (for AIOps incidents) also or do we need to have a separate connector exclusively for ticket creations/ updates and another one for ticket analysis and change risk.

Yes, the same connector can be used. In fact, it's not possible to have more than one ServiceNow connector at a time.

n) Similar to netcool where the alarm is cleared but the service now ticket is not closed, alarm remains in the system until we receive the close status updated back to alarm right. How can we control the same in AIOps?

If an incident is closed in AIOps, it should get closed in ServiceNow. If it doesn't, there is likely an issue occurring. If an incident in ServiceNow is closed AND 2-way sync is enabled, then it should also get closed in AIOps.

o) In case if we need to reopen the service now ticket, how can we pass the updates (severity change, aiops incident status or any other parameters) to service now and will service now consider all these changes while reopening the ticket

In ServiceNow, once an incident is closed, that record is locked and cannot be moved into other states like "in progress". Cancelled/resolved incidents can still be edited and moved into other states. Reopening a servicenow ticket will not reopen the corresponding AIOps incident. Once an incident in AIOps is closed, there is currently no way to re-open it (and it will be deleted entirely shortly after)

p) While two way sync is enabled, if the update is not synchronized back to AIOps, will it try again during next sync or the failed updates will be ignored?

Yes, it will try again on the next resync.

 

Related Information

[{"Line of Business":{"code":"LOB77","label":"Automation Platform"},"Business Unit":{"code":"BU048","label":"IBM Software"},"Product":{"code":"SSE9G0Q","label":"IBM Cloud Pak for AIOps"},"ARM Category":[{"code":"","label":""}],"Platform":[{"code":"","label":"Platform Independent"}],"Version":"All Versions"}]

Product Synonym

AIOps

Document Information

Modified date:
07 September 2026

UID

ibm17277072