CDC without Kafka

What You'll Learn

The clip opens with the proof: one orders table open in two databases side by side. An insert, an update and a delete run against MySQL, and each one shows up in PostgreSQL. The rest of the video builds that stream from scratch.

Tutorial Coverage

• Live proof first: insert, update and delete replicated while both ends are on screen
• Source and target: picking the MySQL database changes come from and where they land
• CDC mode + Initial Load: copy the rows already in the source, then continue from the matching binlog position
• Monitor: snapshot → handoff → live capture, with the handoff position on the card
• Compare: both ends of the run side by side — same schema, same rows
• Verification: writing to the source and watching the row arrive on the target

Prerequisites

• DBConvert Streams deployed and running
• MySQL source with binary logging enabled (ROW format)
• PostgreSQL target — the initial load needs the selected target tables empty
• Network access between source and target databases

Key Benefits of CDC

• Real-time sync: Changes replicated instantly
• Minimal impact: Low overhead on source database
• Incremental updates: Only changed data is transferred
• Resume: Durable checkpoints for controlled restart recovery