We use Microsoft Active Directory. for authentication and authorization. To keep data updated we have a small ETL process in PowerScript that retrive data from our IDM-system and upload it to AD. I have illustrated this by the use of two serving relations. The AD rely on the process and again the IDM system.
That is all fine!
I then add the ADWS-interface, that the ETL proces uses to update data, and because an interface is where a service is made available to another application I ad
the service that is made available and have named it "Update AD data".
But then I get a little confused. Because now the "Update AD data" services is serving the ETL Process and the direction of the serving relations changes direction.
What am I missing - or is it just how Archimate is?
Quote from: AllanV_dk on June 04, 2020, 12:01:31 PM
To keep data updated we have a small ETL process in PowerScript that retrive data from our IDM-system and upload it to AD. I have illustrated this by the use of two serving relations.
Both could exist but mean different things. "ETL Process" serves "Active Directory" is more a functional dependancy (your AD is not up to date without "ETL Process"), while "Active Directory" serves "ETL Process" is more a technical dependancy (to run, your "ETL Process" rely on some web service).
And as your focus is on information, I won't use serving relationships in your "high level view", but flows: one flow from "IDM" to "ETL Process" and another one from "ETL Process" to "Active Directory".
As a side note, you in fact shown that in your context
, "Active Directory" relies on its own "Update AD Data" webservice because if this service fails, AD is still up and running but no more accurate.