Change data capture.

also: CDC · Debezium

Streaming every insert, update and delete from a database's transaction log (WAL or binlog) to other systems as events.

the two-minute refresher

A connector such as Debezium tails the WAL, so it sees every committed change, including deletes.
No dual-write in application code: the database stays the source of truth.
Typical pipeline: PostgreSQL → CDC → Kafka → Elasticsearch or a cache.
Lower latency and load than polling; more moving parts to run.

also: CDC · Debezium

Streaming every insert, update and delete from a database's transaction log (WAL or binlog) to other systems as events.

A connector such as Debezium tails the WAL, so it sees every committed change, including deletes.
No dual-write in application code: the database stays the source of truth.
Typical pipeline: PostgreSQL → CDC → Kafka → Elasticsearch or a cache.
Lower latency and load than polling; more moving parts to run.