Stream Governance is Confluent 17,245 's name for the managed Schema Registry of Schema Registry plus a catalog and lineage built on it. It is attached to an environment (a group of clusters), and its Schema Registry is created in the region of the environment's first cluster and cannot move. Two packages exist:
| Feature | Essentials | Advanced |
|---|---|---|
| Price | Included at no charge | From $1 an hour, plus schemas over the quota |
| Schemas included | 100 (all versions count) | 20,000 |
| Schema Registry SLA | 99.5% | 99.99% |
| Data rules (contracts) | No | Yes |
| Catalog | Tags, REST API, data portal | Adds business metadata, GraphQL API |
| Lineage history | Last 10 minutes | Last 7 days |
The Schema Registry API is the one BookNest used locally, so the Avro 129 schema of Registering Order Schemas registers unchanged and the same serializers find it; only the URL and an API key change. The 100-schema quota counts every version, including schemas that Connect and Flink 129 register for you and soft-deleted versions, so a busy development environment reaches it sooner than the number suggests. Advanced adds data rules: a data contract attaches conditions to a schema, written in Google's Common Expression Language, such as "total is positive" or "email matches a pattern", with an action such as raising an error or sending the record to a dead-letter topic. The serializers enforce the rules on the client, and the dead-letter action works only in the Java client so far: Python, Go, .NET and JavaScript do not support it yet. Data contracts are the streaming counterpart of the ODCS contracts of Data Contracts. Stream Lineage draws producers, topics, connectors and Flink statements as a graph from live traffic, and the Stream Catalog lets you tag topics and fields (for example PII on customers.email) and search them.
Two cautions. Governance features are tied to Confluent's registry: Apicurio Registry (Apicurio Registry) speaks a compatible API for schemas, but tags, rules and lineage do not migrate. And a related product, Tableflow, materializes topics as Apache Iceberg 129 or Delta Lake 237,929 tables in managed or your own object storage, billed per topic-hour and per GB processed; it overlaps the lakehouse ingestion that Streaming into the Lakehouse builds with Kafka Connect 129 , at the price of tying the lake's front door to Confluent.