Debezium Server - New CDC runtime on the horizon!
summary
Debezium Server appears as a standalone CDC runtime that replaces Kafka Connect’s embedded connectors with a three-stage architecture: source capture, parallelized processing, and configurable sinks. Demo deploys the Debezium Operator on OpenShift, points a server at PostgreSQL, shows streaming of table changes (including blobs) into a Kafka-style event format, and reports 539+ contributors for the project. It claims higher throughput by avoiding the single-thread Kafka Connect bottleneck, exposes Kubernetes secrets, and currently guarantees ordered delivery while promising optional unordered modes and pluggable producers.
AI-written summary · about
chapters (12)
- 0:00Introduction to CDC and Debezium
- 0:22Understanding Change Data Capture (CDC)
- 2:07Implementing Poll-Based CDC
- 3:31Challenges with Poll-Based CDC
- 5:40Introduction to Log-Based CDC
- 7:05Challenges with Log-Based CDC
- 8:32Overview of Debezium
- 9:28Debezium Components and Architecture
- 13:21Debezium Server vs Kafka Connect
- 15:17Deploying Debezium on Kubernetes
- 16:14Practical Demo of Debezium
- 21:16Q&A Session
related talks
01h 09mOpen-source Change Data Capture With Debezium
48mStreaming Database Changes with Debezium
42mData Streaming for Microservices using Debezium
42mWhy PostgreSQL people should care about Kafka and Debezium?
31mKafka and Debezium at trivago
41mChange Data Streaming Patterns in Distributed Systems
23mChange Data Capture: A RabbitMQ and Debezium Approach
46mSpell of Transmutation - Using Debezium to Transfer Realtime Data Changes to Event Hub