We used to leverage database for Event Sourcing and CQRS with light-eventuate-4, light-tram-4j and light-saga-4j. Due to multiple dependencies, it is hard to use. With the latest Kafka supports transactions, we have developed Light-Kafka to use Kafka Streams for Event Sourcing and CQRS. When using light-kafka with Event Sourcing, we don’t put producer endpoints and consumer endpoints in the same service. For the rest of the document, we will call the producer as the command service and consumer as the query service. Read More »

Light-4j uses a light-weight HTTP core that can serve millions of requests per second. None of the existing SQL databases can have the throughput to match it. When using the SQL database, we usually use an in-memory cache to serve the data faster. Or we go with Kafka streams for the local RocksDB store for key/value access. For direct database access, please refer to the examples in techempower repo. Read More »