Table of Contents
Status

Current state: ["Under Discussion"]

Discussion thread:

JIRA:

...

Page properties

Please keep the discussion on the mailing list rather than commenting on the wiki (wiki discussions get unwieldy fast).

...

The SchemaRegistryCatalog interacts with the Confluent Schema Registry Service directly through its java SDK client, e.g. the SchemaRegistryClient.

when open/initialize, it fetches all the topic list of current Kafka cluster and makes a cache in the memory.
when a specific catalog table info is requested from the SQL context, it fetches the Avro schema string and format type of the topic's latest subject schema, then generates a catalog table instance and put it into the local cache.
~~We dropped the idea to have table cache because the cache would introduce in-consistency with the latest SR state.~~

A catalog table instance with the following attributes was generated each time #getTable is invoked:

...

The SchemaRegistryCatalog.Builder can be used to configure the following options:

name	isOptional	default	remark
connectorOptions	false	(null)	Connector options for all the tables read from this catalog, the options are shared for all the tables, if you want to tweak or override per-table scope, use the dynamic table options or CREATE TABLE LIKE syntax.
schemaRegistryURL	false	(null)	Schema Registry URL to connect to the registry service.
dbName	true	kafka	database name
schemaRegsitryClient	true	CachedSchemaRegistryClient with default identityMapCapacity of 1000	Sets up the {@link SchemaRegistryClient} to connect to the registry service. By default, the catalog holds a {@link CachedSchemaRegistryClient} with 1000 as {@code identityMapCapacity}. This method is used for custom client configuration, i.e. the SSL configurations or to change the default {@code identityMapCapacity}.

...

Page tree