Before the start

ShannonBase revolutionizes data analytics by natively integrating full machine learning capabilities directly within the database engine, eliminating data movement barriers and enabling intelligent decision-making at the data source.

Zero Data Movement Architecture ensures training and inference occur within transaction boundaries, removing ETL overhead and enabling real-time feature engineering directly at the storage layer.

The design philosophy of ShannonBase Rapid is modularity and performance-cost balance. The following outlines the new features that will be implemented in ShannonBase. To learn details about each feature, see the relevant chapter. ShannonBase Rapid will still be an open source project, which is a counterpart of close source service, MySQL Heatwave. At first, an in-memory column store (IMCS) will be used. Secondly, a cost-based query engine will be developed to automatically offload transactional and analytics workloads. Thirdly, ShannonBase Rapid will provide a vectorized execution engine and support massive parallel processing. In this way, the execution performance of ShannonBase Rapid will be at least xxx times as that of xxx. ShannonBase will load the data into memory from InnoDB to Rapid, just the same as MySQL Heatwave does. ’‘’MySQL Analytics is an in-memory processing engine, data is only persisted in MySQL InnoDB storage engine.‘’‘ This sentence functions as the basic rule and guideline for us when implementation ShannonBase Rapid. This design document introduces the main changes that will be achieved and gives you an overview of architecture of ShannonBase.