Guide
Create Systems and Functions
Model the service areas you actually route, report on, and work during incidents.
Before you start
- Pick the service boundaries people already use in incidents.
- Decide which items should be top-level systems and which should be child functions.
Do this
- Open
Catalogand create each top-level system you want responders to recognize. - Add a short description for each system so people know what it covers.
- Create functions for the parts of the system that change ownership, alerting, or impact separately.
- Attach each function to the correct parent system when that relationship exists.
- Keep names stable and human-readable so alerts, incidents, and reports all use the same language.
- Edit the model when ownership or architecture changes instead of adding near-duplicate items.
Check it worked
- The catalog reflects how teams already talk about the service.
- Functions sit under the right systems.
- The model is specific enough to route alerts and broad enough to stay maintainable.
If it does not work
- If the model feels too deep, move detail out of the catalog and keep only the routing-relevant layers.
- If teams keep picking the wrong target, rename the objects before adding more.
- If ownership is unclear, pause new catalog work until that is agreed.
