As part of our ingestion strategy we want, for certain properties, to ingest their data from multiple sources. Usually one that will ingest inferred data and another that will ingest declared data. e.g. Service Team Name can be inferred from the k8s deployment's metadata so we'll be able to populate this property for all Service entities but for such services where the inference is incorrect, we want to be able to override the inferred team name with a declared team name, e.g. using a metadata yaml file stored in the service repository.

Today, if we'll configure ingestion from multiple sources, there's no ordering guarantees by which ingested data will take precedence over the other. Having an ability to define each data source priority will allow us to easily ingest data in bulk on the one hand and still allow for declarative ingestion and manual override for the cases where inference is incorrect.