Achieve a high-speed InnoDB purge on Amazon RDS for MySQL and Amazon Aurora MySQL

This article discusses how to achieve a high-speed InnoDB purge on Amazon RDS for MySQL and Amazon Aurora MySQL databases.

Specifically, the article covers:

• Understanding how the InnoDB purge process works, including undo logs, rollback segments, and the purge coordinator/worker threads
• Balancing read and write workloads to avoid blocking purge due to long-running queries or transactions
• Optimizing table and index structures to reduce overhead for purge worker threads
• Choosing the right instance class with adequate CPU, memory, and I/O resources for efficient purge operations
• Monitoring and setting alarms for the rollback segment history list length metric to detect purge lag