The Operator Feature Matrix (OFM) is a project from the Data on Kubernetes (DoK) Community to create standardized and vendor-neutral feature matrix for various Kubernetes operators that manage stateful workloads. The key concepts behind the project are:
- The Feature Matrix. A categorized and structured set of features that a given operator (for a particular technology) may support.
- Standardized features. Define terminology for features so that naming is standardized.
- Vendor neutral. OFM is a project started by the DoK with contributions from various organizations and contributors.
The main goals of the OFM are:
- Allow end users to quickly identify if a given operator supports or not a given feature or set of features.
- Provide clear and standardize feature naming to avoid end users being confused about the meaning of particular features provided by an operator.
- Enable end users to produce objective, side-by-side comparisons of the features implemented by different operators (for a given technology).
The OFM is developed by consensus, and standardized; submissions of the features implemented by each vendor at a particular operator release are expected to be contributed by the vendors themselves or their respective Communities.
The OFM is technology-dependent, as most of the features of a given technology (e.g. the Postgres relational database) are different from the features end users expect from another different technology (e.g. Kafka). As of today, there is one OFM implemented:
It is expected that other technologies may add their respective OFM to this project, following similar directions to the one already existing. However, it is also expected that once a few technologies and their respective OFM are added, a common subset of features that may span across all potential technologies may be identified. In this case, a cross-technology OFM subset could be created, and existing projects (partially) retrofitted to that model, so that technology-independent features are identified as such.