dataarchitect.studio

Glossary

Change Data Capture (CDC)

Change data capture (CDC) is the technique of extracting changes from a database — every insert, update, and delete — by reading the transaction log the database already writes, rather than polling tables with timestamp filters. The log is ground truth: changes come out complete, ordered, and including deletes, with near-zero load on the source.

CDC is how operational data reaches analytical systems with minutes-level freshness, and log-based CDC (e.g. Debezium) is the standard modern implementation. The capture itself is the easy half; applying the stream to a target correctly requires an idempotent merge, schema-change handling, and an initial snapshot — the operational machinery around the log reader.

Go deeper: What Is Change Data Capture (CDC), and When Do You Need It?