TiCDC background and positioning (Legacy)

TiCDC is TiDB’s change data capture (CDC) component. It streams changes from TiKV to downstream systems (MySQL / Kafka / storage / etc.) and is commonly used for:

• Disaster recovery (DR) / replica clusters
• Data integration (streaming changes into search, analytics, and real-time pipelines)

1. Why TiCDC (and not TiDB Binlog)

Historically, TiDB had TiDB Binlog as a replication toolchain. TiCDC was introduced to address major limitations commonly seen in TiDB Binlog deployments, such as:

• Limited scalability (e.g. single-node bottlenecks)
• Throughput constraints in common MQ setups
• Protocol/ecosystem constraints (harder to integrate with standard CDC consumers)
• Operational fragility and limited self-healing behavior

In short: TiCDC is designed as a more scalable, more maintainable CDC pipeline for modern integration/DR use cases.

2. Where TiCDC sits in the TiDB ecosystem

You can think of TiCDC as the “egress” layer between TiDB and downstream systems. In practice, downstream targets include:

• MySQL-compatible databases (TiDB / MySQL / RDS-like services)
• Message queues (Kafka, etc.) for stream processing (Flink / Spark / custom consumers)
• Custom sinks via supported protocols (Open Protocol / Canal / Maxwell / Avro, etc., depending on version/support level)

3. Typical pipelines

Common production patterns:

• TiDB → TiCDC → Kafka → Flink → (Warehouse / Serving / TiDB)
• TiDB → TiCDC → MySQL (for a replica / validation / cutover)
• TiDB → TiCDC → MQ / sink for search, recommendation, and real-time analytics

4. Getting started (recommended path)

This page is a legacy note. For the most accurate and up-to-date configuration, use the official docs:

• TiCDC overview: https://docs.pingcap.com/tidb/stable/ticdc-overview
• Scaling TiCDC using TiUP: https://docs.pingcap.com/tidb/stable/scale-tidb-using-tiup#scale-out-ticdc-nodes

If you want a quick sanity check, validate a minimal changefeed to:

• Kafka (validate topic/partitioning and message format)
• MySQL (validate DML/DDL propagation and lag)

References

• Infra Meetup: Why and how we build TiCDC: https://www.bilibili.com/video/BV1HT4y1A7mh
• Infra Meetup: TiCDC ecosystem and community: https://www.bilibili.com/video/BV1bD4y1o7qU
• TiDB Binlog overview: https://docs.pingcap.com/tidb/stable/tidb-binlog-overview
• TiCDC protocol list (design doc): https://github.com/pingcap/tiflow/blob/master/docs/design/2020-11-04-ticdc-protocol-list.md