Building Kafka-Less Data Integration Pipelines with Debezium Navigation Menu<br>Home
Archive
Feed
Featured Posts<br>Debezium Platform: Mid-Year Update (2026)<br>Debezium Community Showcase<br>What Nobody Explains About Debezium in 2026 (But Should)<br>Join the Debezium Community Weekly Forum<br>Detect data mutation patterns with Debezium<br>Debezium asynchronous engine<br>Debezium and TimescaleDB<br>Streamlined Performance: Debezium JDBC connector batch support<br>Debezium Operator Takes off to the Clouds<br>Debezium signaling and notifications - Part 3: JMX channel<br>Online machine learning with the data streams from the database<br>Debezium signaling and notifications - Part 2: Customisation<br>Debezium signaling and notifications - Part 1<br>Image classification with Debezium and TensorFlow<br>Streaming Vitess at Bolt<br>Distributed Data for Microservices — Event Sourcing vs. Change Data Capture<br>Building Audit Logs with Change Data Capture and Stream Processing<br>Streaming Cassandra at WePay - Part 1<br>Reliable Microservices Data Exchange With the Outbox Pattern<br>Automating Cache Invalidation With Change Data Capture<br>Materializing Aggregate Views With Hibernate and Debezium<br>Five Advantages of Log-Based Change Data Capture<br>Creating DDD aggregates with Debezium and Kafka Streams<br>Streaming Data Changes from Your Database to Elasticsearch
Tags<br>UI WebAssembly ai analytics announcement announcement discussion survey antlr apache kafka apache kafka apicurio avro aws batch caassandra camel cassandra cdc cdc tables channels charts chicory cockroachdb commonhaus community community stories configuration connectors containers contributions contributors core cqrs custom data data engineering data replication datalake db2 ddd ddl debezium debezium operator debezium platform debezium server debezium ui debugging deduplication demo design patterns discussion dlt dlthub docker elasticsearch embedded engine event sourcing events exactly once semantics example examples features fedora flamegraphs flink forum go graalvm gsoc hiring howto ibmi iceberg images informix integration internals introduction issues jaeger jdbc json jupyter kafka kafka streams kafka streams kogito ksql kubernetes lakehouse lineage llm logminer machine learning mandrel mariadb meetings metrics microservices migration mongo mongodb monitoring mysql native news newsletter notifications online learning openlineage operator oracle outbox pandas parsing performance platform postgres presentation production pydbzengine python quarkus questdb rag rds refactoring releases replication schema scylla secrets sentry serialization showcase signaling smt snapshots spanner spark sql sqlserver streaming tensorflow testcontainers tests time series timescaledb tinygo topics tracing transactions troubleshooting tx log ui update vagrant vitess wasm website yashandb
Building Kafka-Less Data Integration Pipelines with Debezium Building Kafka-Less Data Integration Pipelines with Debezium July 6, 2026 by Fiore Mario Vitale debezium-server debezium-platform replication migration
Debezium is often associated with Kafka, and while this was true in the project’s early days, it is no longer the case. We explained in detail, together with several other common misconceptions, in our recent article, "What Nobody Explains About Debezium in 2026 (But Should)".
In this article, we’ll build on that discussion by showing the recently added JDBC sink support in Debezium Server (DS) and the Debezium Management Platform (DMP), which opens up Kafka-less data replication and migration scenarios.
Data integration, whether for one-shot migrations or continuous replication between databases, is one of the core use cases Debezium supports. Debezium has always focused on the CDC side of the problem: capturing changes from a database and turning them into a stream of change events. In the past, users had to develop their own consumer to write those events into a target database. To address this, we introduced the JDBC Sink Connector in the 2.2.0.Beta1 release in 2023, enabling end-to-end replication out of the box. The primary goal was not just to provide another database sink, but to provide full end-to-end support for Debezium’s native change event format, including source metadata and change semantics, without requiring event transformation or loss of context. At the time, this level of support for Debezium’s native events was not available in traditional relational database consumers.
Today, the same capability is available through the DS JDBC sink and within the DMP. This is not intended to replace Kafka Connect (KC) connectors. Instead, it provides an alternative for the users who do not have a KC cluster, or who do not need the scalability and high availability that KC offers, and simply want to move data between two databases. In those cases, DS or DMP can now be an attractive option, and with the upcoming native build support, DS will be able to run as a single native executable with minimal resource footprint.
All the examples...