Evolving data products
Data products have a life cycle and undergo continuous evolution. This section delves into strategies for managing the evolution of data products while minimizing the impact on their consumers.
Versioning data products
All consumers of a data product access the data and, more generally, the services it offers only through the APIs it exposes through its ports. Therefore, managing the life cycle and evolutions of a data product is not particularly different from managing the life cycle of an API.
Each data product can have multiple versions, but in a given runtime environment, only one instance of the data product per version can exist. The instance with the highest version number present in the production environment is the current version of the data product. The versioning of a data product must use semantic versioning (https://semver.org/). According to semantic versioning specification, a version must be defined using three numbers: major, minor...
