Schema and Index Optimization

A well-designed table schema, the right index choices, and efficient scan strategies determine the upper bound of query performance. Optimize layer by layer in the following order.

Table Schema Design Optimization

Choose the table model, partitioning and bucketing columns, Key columns, and field types to avoid data skew and fully use parallelism.

Table Index Design Optimization

Pick the prefix index, BloomFilter, NGram BloomFilter, or inverted index based on query patterns, and apply them to the right columns.

Table Scan Optimization

Use partition pruning, prefix index pushdown, and Runtime Filter to scan less data on every query.