DBRaven
Relationship · Supports
Source: Technology·Target: Pattern

Summary

Kafka is the standard downstream target for WAL-based CDC pipelines: Debezium captures database WAL records and publishes them to Kafka topics, which downstream consumers process to maintain derived data stores, caches, and event-driven services.

Evidence

  • ·Debezium (Red Hat) captures PostgreSQL, MySQL, and MongoDB WAL and publishes to Kafka topics
  • ·Confluent Platform bundles Kafka Connect with Debezium connectors as a first-class CDC product
  • ·LinkedIn's Databus (2013) was an early WAL-to-Kafka CDC system that inspired the Debezium architecture
  • ·Kafka's log compaction enables CDC topics to serve as full-state snapshots for new consumers
  • ·Netflix uses WAL-to-Kafka CDC to maintain Elasticsearch indexes in sync with their primary PostgreSQL databases

Operational Context

  • ·Debezium replication slot on PostgreSQL must be monitored: a lagging or disconnected Debezium causes replication slot WAL accumulation on the primary
  • ·{'Kafka topic naming convention for CDC': 'dbname.schemaname.tablename (Debezium default)'}
  • ·Consumer groups on CDC topics must handle schema evolution: Confluent Schema Registry with AVRO provides compatibility

Tradeoffs

  • ·Debezium replication slot holds WAL until consumed: disconnected Debezium can fill primary disk
  • ·CDC events are row-level operations: application-level event semantics require transformation in a stream processor
  • ·Ordering guarantees are per-partition only: cross-table ordering requires careful partition key strategy

Evidence grounding

Grounded, 5 supporting items

The Debezium + Kafka CDC pattern is the de facto standard for CDC-to-streaming, documented and used in production at hundreds of companies. Kafka's durability and consumer group model makes it the natural CDC sink.