KafkaAvroSerializer properties reference
Review the following reference for a comprehensive list of the KafkaAvroSerializer properties.
| Property Name | Description | Type | Default Value |
|---|---|---|---|
| schema.compatibility | The default compatibility of the new schemas created by the serializer. For more information about the compatibility options, see the Compatibility policies documentation. | String | BACKWARD |
| schema.group | The default group of the new schemas created by the serializer. | String | kafka |
| schema.name.key.suffix | The provided string will be appended to the end of the schema name created for record keys. The schema name is the topic name initially, so as a result, you will have topic name + provided suffix as schema name. | String | :k |
| schema.name.value.suffix | The provided string will be appended to the end of the schema name created for record values. The schema name is the topic name initially, so as a result, you will have topic name + provided suffix as schema name. | String | null |
| null.passthrough.enabled | If this is enabled, the serializer will return immediately with null value, without putting anything in the serialized output. Having null values can be important in compact topics. Having null as the latest value for a key indicates for Kafka that the corresponding key can be deleted during log compaction. | Boolean | false |
| logical.type.conversion.enabled | If this is enabled, Avro record fields can be serialized that are instances of
Java classes (BigDecimal, UUID, LocalDate, and so on) that can be serialized in Avro
using logical types. For more information, see the Logical Types section in the official Avro
documentation. This does not have to be enabled if SpecificRecords are used as part of generated classes. SpecificDatumWriter is used for serializing SpecificRecord instances and that writer handles logical types automatically. |
Boolean | false |
| serdes.protocol.version | The protocol version to be used. The available default registered protocols are
the following:
|
Byte | 0x3 (VERSION_ID_AS_INT_PROTOCOL) |
| store.schema.version.id.in.header | If this is set to true, the schema version ID will be stored in the record header instead of the value. | Boolean | false |
| key_schema_version_id_header_name | If the schema of the record key is stored in the header, the header name can be configured. | String | key.schema.version.id |
| value_schema_version_id_header_name | If the schema of the record value is stored in the header, the header name can be configured. | String | value.schema.version.id |
The KafkaAvroSerializer contains an embedded
SchemaRegistryClient. The configuration for
SchemaRegistryClient should also be provided for the
KafkaAvroSerializer. For more information about configurable properties,
see the Schema Registry Client properties reference documentation.
