I'm proposing the use of building B2B relationships for BSO to BSO in an effort to limit footprint/impact in changes and incidents.
This is what CSDM shows:
It shows Application Service to Application Service which does ultimately show the impact of the BSO to BSO but the impact greater at the application service level. Showing BSO to BSO will show that only a subset of functionality of an application is impacted and only the related subset(s) of application functionality of others is impacted. In neither case is the whole application impacting all of the other applications that may depend on it.
If Application ABC has functions 1, 2 and 3 and Application XYZ has functions 7, 8 and 9 and Application LMN has functions 4, 5 and 6 and ABC2 sends data to XYZ7 and XYZ8 depends on LMN6, having B2B at the application service level will show that if ABC has a problem, XYZ, LMN and whatever depends on LMN will have a problem. However, if you have the B2B at the BSO it will show that ABC2 has a problem so XYZ7 has or may have a problem and any BSO that interacts with that may be impacted but LMN is not impacted.
CSDM doesn't call this out to my knowledge but I think this reduces impact but will require more detailed relationship building. Rather than using an application service as the impacted application, you could use the BSO. You can still call out the specific piece of infrastructure that ties to that piece of functionality of that application service which will still allow you to report on that application service by way of it's piece of infrastructure but it will give you a clearer view of which functionality of that application service is impacted in a more "human"/less technical way.
Am I looking at this wrong? Is that causing too much unnecessary work?
