Best Practices for Application Service Management in CSDM

[miyu]

We are trying to build the management of the services we currently operate in compliance with CSDM.
For example, we have an application called "ID Card Management System".
We are thinking of managing the "ID Card Management System" application through an application service.
However, the "ID card management system" application has an application manager and an infrastructure manager.
Therefore, when we raise a defect report incident, we would like to be able to distinguish whether the defect is on the application side or the infrastructure side.
How should we manage the separation of application and infrastructure for the same application service?
What is the best practice?

[Ravi Peddineni]

Are these Application Manager and Infrastructure Manager different application or just infrastructure supporting ID Card Management system? 

If they are different applications, you should have 3 application services:

• ID Card Managemet
• Infrastructure Manager
• Application Manager

Now, Infrastructure and Application Manager application services would support ID Card Management(Relationship) in this scenario.

If they are just Infrastructure in the company, you can just select one of it in the Incident CMDB CI field.

[miyu]

@Ravi Peddineni 

Thanks for the reply.
Application Manager and Infrastructure Manager are the administrators that support the ID Card Management system.
Application Manager supports the application area of the ID Card Management System; Infrastructure Manager supports the infrastructure area or CI of the ID Card Management System.
My supervisor told me that this could be accomplished by using ServiceOffering, but I can't picture it.
Is there a better way to do this?

[Barry Kant]

hi Miyu,

the best practice is that you also model that infrastructure in the model. Downstream of the application service(s). 
If the incident is raised by consumers/customers. then they don't know that infrastructure most of the times, so it starts from an application point of view. If the incident is created by monitoring then it can include the infrastructure object. This object contains the infrastructure party as it is related via dynamic CI group to a Technical Service Offering and the Support Group is synced from that TSO. 
That is about the 80% rule for application solutions. (Not 100% sure if that is the case). 

Or is this system including ID Scanning devices?
Those potentially can be modeled as application services as well (not sure if that is needed) and then a consumer/customer does know the device level. in that scenario the devices includes a client/app and is connected to a centralized application (separate application service). This scenario could work for example for ticket machines where it is interesting to understand the sales capacity on a certain location. 
BR,
Barry

[Diego Peukert]

Hi Miyu,

I've been working with ITOM for some time and have a lot of dependence on CMDB and its modeling, including using CSDM to extract the best from OOTB that cross several modules of the tool, which depend on CMDB.

Regardless of how you are going to build your Application Service, either manually or from Tags or Top-Down (more automated ways) using ITOM-Service Mapping, I would follow in the following order:

1-Class: Application Service
- Support Group = Support Group responsible for the Application, to which the group's Manager is the 'Application Manager
- Owned by - Name of the user responsible for the ID Card Management System

2-Class: Server or other relevant Infrastructure Items that are related in the Application Service - ID Card Management System

- Support Group = Support group responsible for the Infrastructure, for which the Manager of the group is the Infrastructure Manager

I do not recommend using Owned by or Assigned to for these Datacenter Infrastructure Items.

For both classes you can create logic and apply Business Rules to automatically fill these attributes.

Given the customer's IT service management scenario and whether you have only ITSM or ITOM+ITSM, there are different strategies to leverage the modeling I am suggesting.

I agree with Barry's answer, where the solution he reported does not meet 100% of the cases. When more information of the scenario that the customer finds himself in IT service management and modules he has contracted, we can target a better modeling.

Some questions that can better direct the construction:

1 - Does the client have the ITOM + ITSM modules ?
2 - Does the client create his Application Services using which method ?
3 - Has the client adhered to Business/Technical Services > Service Offering > Application Service > Infrastructure/Application modeling
4 - What types of Incidents are opened automatically ?
5 - How do IT users open incidents ?
6 - How does IT support or IT service management open incidents manually ?
7 - Are Infrastructure and Application items monitored ?
8 - Is there automatic opening of Incidents for Infrastructure or Applications ? Can monitoring determine the affected resource and pass it on to SNow ?