Introduction
ModelOp Center provides the ability for a Model Life Cycle to create and manage tickets within an existing ServiceNow system. The ticket types supported include change requests and incidents created and managed using the Table or Import Set APIs. This guide details the steps to enable integration.
ServiceNow Integration Setup
Prerequisites
This guide assumes moc-builder
has been installed. Please request download instructions for moc-builder
here. Before configuring moc-builder
, the ServiceNow instance will need a service account created and configured. For an example of how to set up a service account see the instructions at https://community.servicenow.com/community?id=community_blog&sys_id=b4fca2a5dbd0dbc01dcaf3231f961900. This guide assumes that the service account uses the email "servicenow@modelop.com". There is no need to provide a ServiceNow related flag when bringing up the fleet as the ServiceNow instance will always be an external service.
moc-builder ServiceNow Configuration
The integration requires the ModelOp Center Gateway url and the url of the ServiceNow instance to be integrated.
First, determine the external-ip of the Gateway. Set the Kubernetes context to the cluster and namespace that has ModelOp Center deployed. Run the following command:
kubectl get svc
The command will return an output like this:
[johncarter@Johns-MacBook-Pro moc-builder (John@modelop-eks-test.us-east-2.eksctl.io:mocaasin)]$ kb get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE engine-1 ClusterIP 10.100.72.22 <none> 8003/TCP 23h engine-2 ClusterIP 10.100.31.255 <none> 8003/TCP 23h engine-test ClusterIP 10.100.215.6 <none> 8003/TCP 23h gateway LoadBalancer 10.100.126.106 foo1.us-east-2.elb.amazonaws.com 8090:31166/TCP 23h jupyter ClusterIP 10.100.242.107 <none> 8888/TCP 23h kafka ClusterIP 10.100.113.242 <none> 9092/TCP 23h minio ClusterIP 10.100.82.143 <none> 9000/TCP 23h mlc-manager LoadBalancer 10.100.145.82 foo3.us-east-2.elb.amazonaws.com 8085:31741/TCP 23h model-manager ClusterIP 10.100.205.30 <none> 8086/TCP,8088/TCP 23h mongodb ClusterIP 10.100.45.49 <none> 27017/TCP 23h postgres ClusterIP 10.100.77.22 <none> 5432/TCP 23h registry ClusterIP 10.100.138.128 <none> 8761/TCP 23h reporting-service ClusterIP 10.100.38.87 <none> 8091/TCP 23h
The important value is the external-ip and port corresponding to the ModelOp Center Gateway. In this example, it’s foo1.us-east-2.elb.amazonaws.com:8090.
NOTE: If the environment is using ingresses instead of load balancers, run kubectl get ing
and obtain the address corresponding to the ingress that points to gateway. Do not append the port number to the end of the address. That is only required if you are using the external IP from a service of type LoadBalancer.
Next, open the application.yaml
file within the moc-builder
environment directory and navigate to the config.moc-services.gateway
section. It will contain something like the following:
gateway: url: http://gateway:8090 websocket-url: ws://gateway:8090 external-url: http://localhost:8090 response-timeout: 2m
Change the value of external-url
to the location of gateway from above.
Here is an example of that section of application.yaml
:
gateway: url: http://gateway:8090 websocket-url: ws://gateway:8090 external-url: foo1.us-east-2.elb.amazonaws.com:8090 response-timeout: 2m
Find the config.moc-services.supporting-services.servicenow
section. It will contain something like the following:
servicenow: username: demo password: modelop url: http://servicenow
Change the username
to the username associated with the service account and the password
to be the service account password. Update the url to be ServiceNow url. For this example we use https://dev.servicenow.com
.
Here is an example:
servicenow: username: servicenow@modelop.com password: service account password goes here url: http://dev.servicenow.com
Next, run the moc-builder
compose and deploy commands to bring up the instance of ModelOp Center integrated with ServiceNow.
Security
As mentioned in Prerequisites we recommend using a service account for the ServiceNow integration. Limit the access rights of the service account to just those needed for the integration and only use the service account for the integration. The security advantages to using a service account are explained here.
The integration uses Basic Authentication to authenticate REST calls to ServiceNow. The credentials can be set up using Kubernetes secrets, as in the configuration example above, or using Vault.
ServiceNow Ticket Customization
The integration provides a variety of parameters to allow customization of ServiceNow ticket creation and lifecycle management within the Model Life Cycle. Some of these parameters are discussed in more detail below.
Ticket Creation
SERVICENOW_API
- The API to use when creating tickets (table or import). The integration supports both the Table API and the Import Set API for ticket creation.
SERVICENOW_TABLE_NAME
- The table to create new tickets in. This is the path parameter of the API being used. This is usually change_request, incident, or the name of an Import Set.
SERVICENOW_CUSTOM_FIELDS
- A JSON string containing any custom fields to send with the creation request. By default the implementation will provide data for a few fields and this parameter allows setting as many fields as needed. The custom fields can also be used to override the value traditionally sent for that field. One exception to the override functionality is the description
field. If the description
field is overridden, the custom override value will be used as a notification message within the description and the default description template will still be used to ensure that links back to the appropriate entity are included in the description. If a uuid
or u_uuid
field is required per request, including retries, then set the field to empty string and the integration will generate a random UUID to use for each request.
Ticket Updates
SERVICENOW_UPDATES
- A JSON string specifying additional updates to make to the ticket after creation. Some workflows require additional automated updates to the ticket after creation. This field includes the required information to make those updates. UUID generation works here the same as for SERVICENOW_CUSTOM_FIELDS
. If a field needs to use the value from the current state of the ticket include that field in the payload and set the value to empty string. The integration will replace the value with the value
of the field in the ticket. The structure of the JSON is shown below:
[ { "serviceNowApi" : "(table|import)", "serviceNowTableName" : "tableName", "serviceNowCustomFields" : { "field1":"value1", "replacement": "", ... } ... ]
Ticket Lifecycle
SERVICENOW_EXIT_STATUS_FIELD
- The field on the ticket to watch for an exit state. The Model Life Cycle waits until this field reaches a desired SERVICENOW_EXIT_STATUS
. By default the integration watches the state
field. The display_value
of the field is watched.
SERVICENOW_EXIT_STATUS
- A string listing out the exit statuses for the ticket. When the SERVICENOW_EXIT_STATUS_FIELD
reaches one of the exit statuses in this parameter the Model Life Cycle will stop watching the ticket and proceed.
SERVICENOW_EXIT_CONDITIONS
- A string listing out a list of statuses and exit conditions that must be met for the status to be realized. Some workflows require considering multiple fields to determine if the ticket has met its exit status. This field uses a CSV-like format to list out conditions. Specifically,
Status1,ConditionType1,Condition1 Status2,ConditionType2,Condition2 …
where commas are used as a separator. Commas are permitted in the condition. A new line indicates the beginning of a new condition. Both juel and spel conditions are supported.
As an example, a workflow may require watching the state
field to reach canceled
or for the approval
field to reach approved
or rejected
. The conditions for such a workflow could look like the following.
Canceled,juel,\${ticket.state.display_value.equalsIgnoreCase('canceled')} Rejected,juel,\${ticket.approval.display_value.equalsIgnoreCase('rejected')} Approved,juel,\${ticket.approval.display_value.equalsIgnoreCase('approved')}
The conditions are evaluated in the order they are listed. The ticket
variable is provided for expression evaluation by the integration and represents the current state of the ticket. The \$
is escaped in the example above so that Camunda itself will not attempt to evaluate the juel conditions. When a condition is satisfied the TICKET_STATUS
variable will be set to the matching status for further use in the Model Life Cycle.
Only one of SERVICENOW_EXIT_CONDITIONS
and SERVICENOW_EXIT_STATUS
will be used when determining if a ticket has reached exit status with the following priority order:
Local
SERVICENOW_EXIT_CONDITIONS
variableLocal
SERVICENOW_EXIT_STATUS
variableSERVICENOW_EXIT_CONDITIONS
variableSERVICENOW_EXIT_STATUS
variable
In cases 2 and 4 SERVICENOW_EXIT_STATUS
can be set to null to indicate that there should be no wait after creation and update. If SERVICENOW_EXIT_CONDITIONS
is used then SERVICENOW_EXIT_STATUS_FIELD
has no affect.
Raising/Managing Change Requests
To use a change request within a Model Life Cycle a notification must be created and stored and then the ticket created and the lifecycle managed. We provide examples of how to use the different notifications in the mlc-building-blocks. A simple example is shown below.
The Simple Notification requires the following parameters.
NOTIFICATION_MESSAGE
- A message to include in the notification. This message will appear on the ModelOp dashboard. It will also appear as part of the short_description
and in some cases description
of the ServiceNow ticket unless those fields are overridden using SERVICENOW_CUSTOM_FIELDS
.
NOTIFICATION_SEVERITY
- The severity of the notification (INFO, WARN, ERROR, etc.). This will be reflected in the ModelOp dashboard notification.
SERVICENOW_API
- The API to use for creating the ServiceNow ticket (table or import).
SERVICENOW_TABLE_NAME
- The ServiceNow table to create the ticket in. If using the Table API this will be change_request. If using the Import Set API this will be the name of the import set.
Supplying these parameters correctly will result in a dashboard notification and a ServiceNow ticket being created. After ticket creation, however, the Model Life Cycle will continue on and not wait for any specific status of the ticket. In some cases this is desirable but often waiting for a specific status is the desired behavior. The SERVICENOW_EXIT_STATUS
variable can be used to set the the status to wait for. For example, if set to “Closed,Canceled” then the Model Life Cycle will wait until the ticket’s state.display_value
has reached either the closed or canceled status before continuing. The final status will be set in the TICKET_STATUS
variable which the Model Life Cycle can use to determine different paths of execution. See ServiceNow Ticket Customization for more parameters for customizing ticket creation, updates, and lifecycle management.
Raising/Managing Incidents
Raising and managing an incident is nearly identical to Raising/Managing Change Requests. Ensure that SERVICENOW_TABLE_NAME
is set to either incident (if using the Table API) or the name of the appropriate import set. That’s it! An incident will be created instead of a change request. Update any other parameters as needed to manage creation, updates, and lifecycle.
MLC Example Usages
The most common pattern when managing ServiceNow tickets in a Model Life Cycle is a three step pattern of Create Notification, Store Notification, and Review Notification. This pattern occurs twice in the RunAssociatedModelMetricsServiceNow
Model Life Cycle as shown below. In each of the three step patterns the first activity is used to indicate the type of Notification to create as well as the ServiceNow notification assignment information. (See mlc-building-blocks for more examples of creating ServiceNow tickets.)
Each notification created will appear on the notification dashboard (Model Review Notifications also appear on the snapshot page). After the ticket is created the notification will be updated with a link to the ticket. The ticket status for the notification will be updated periodically while the Model Life Cycle is waiting on the exit status of that ticket. An example of one of these dashboard notifications is shown below.