Connect Data
CDC
Change Data Capture streams row-level inserts, updates, and deletes from a source database in real time, instead of relying on periodic full reloads.
Quick Start
- 1
Choose a source database type
Pick the database engine you want to capture changes from — PostgreSQL, MySQL, Oracle, SQL Server, or MongoDB — or select an existing data source already configured under Data Sources.
- 2
Set up the connection
Provide connection details for the source database, the same as configuring any database source.
- 3
Select tables
Choose which tables to capture changes from — you don't have to capture the whole database.
- 4
Configure the connector
Set any connector-specific options for the CDC job.
- 5
Review & deploy
Confirm the setup and deploy the connector. It starts streaming changes as soon as it's live.
Source database engines
Each engine uses Debezium under the hood, but requires a specific setting to be enabled on the source database before CDC can run:
PostgreSQL
RecommendedLogical replication via Debezium. Requires wal_level=logical.
MySQL
Binlog-based CDC via Debezium. Requires binlog_format=ROW.
Oracle
LogMiner-based CDC via Debezium. Requires ARCHIVELOG mode + supplemental logging.
SQL Server
CDC via Debezium. Requires SQL Server CDC enabled + SQL Server Agent running.
MongoDB
Change streams via Debezium. Requires a replica set or sharded cluster.
- You can reuse an existing, already-configured database source instead of setting up a new connection from scratch.
- PostgreSQL is the recommended engine — it has the most mature logical replication support via Debezium.
- Selecting fewer tables at step 3 reduces load on the source database and speeds up initial setup.