So we can capture more information for monitoring, such as better identify which services are slow to startup.
Currently the events are a bit coarse grained. We need fine grained for
- per route
- per service
And maybe have a way of grouping/categorizing services so you can better understand what they are used for.