Competition¶
Schema Registry is only one of the types of registries this project supports. On the same foundation, other registries such as MLRegistry, Rules, ApplicationConfigs etc. can be built. The foundation offers StorageManager, REST API structure and a Versioned Entity store upon which other modules are built.
Specifically in SchemaRegistry, we are providing a solution that can capture various formats of Schema and provide pluggable serializer/deserializers for schemas.
Any clients such as Storm, Spark, Nifi, Kafka and other projects can easily use the schema-registry-client to query schema registry to serialize/deserialize the messages.
A frequently asked question is, how Schema Registry from this project is different from Confluent Schema Registry. While we don’t have a clear roadmap of Confluent Schema Registry, here are the current capabilities of both. You may want to look at future roadmap.
Features | Registry | Confluent Schema Registry |
---|---|---|
REST API for
Schema Management
|
Yes
|
Yes |
Supported Schema Types
|
AVRO (0.3), PROTOBUF,
JSON, extensible…
|
AVRO only |
Pluggable
Serializer/Deserializer
|
Yes
|
No, uses Avro parser |
Storage
|
Pluggable Storage
(Mysql, Postgres)
|
Uses Kafka as storage |
Registry Client to support
multi platform(Storm,Kafka
, Nifi etc..)
|
Registry client to
interact with schema
registry
|
No. Developed for Kafka producer/consumer APIs only |
HA
|
No Master,Multi
Webserver deployment
|
Single Master Architecture Depends on zookeeper |
UI
|
Yes
|
No
|
Collaboration
|
In future- See Roadmap
|
? |
Search
|
Yes
|
No |