Two Approaches to Synchronize MySQL Data with Redis Cache

This article introduces two solutions for synchronizing MySQL data to a Redis cache.

Solution 1: MySQL Trigger + UDF

When data is modified in MySQL, a trigger fires and calls a user‑defined function (UDF) that writes the changed data directly to Redis, achieving real‑time sync. This method suits read‑heavy, low‑write scenarios but can degrade performance on frequently updated tables.

Solution 2: Binlog Parsing

By parsing MySQL binary logs, the system extracts data changes and writes them to Redis, similar to MySQL replication where the master writes to binlog and the slave reads it. This approach works for any write pattern but requires deep understanding of binlog formats and handling of statement/row/mixed modes.

Both solutions are illustrated with diagrams and example configurations, including trigger definitions and UDF code snippets (shown as images).

Canal Open‑Source Project

Canal, an Alibaba open‑source Java project, simulates a MySQL slave to receive binlog streams, parses them, and provides a pipeline (eventParser, eventSink, eventStore, metaManager) for further processing. By extending the sink or store, parsed data can be written to Redis.

Additional notes: The article warns against a reverse approach (write to Redis first then sync to MySQL) due to reliability concerns, and provides references to related open‑source projects and further reading.