Google Cloud created BigQuery history-based optimizations, a novel query optimization method that uses past executions of related queries to find and implement further query execution enhancements
Google observed cases where query speed on customer workloads increased by up to 100x during the public preview of BigQuery history-based optimization
Google observed cases where query speed on customer workloads increased by up to 100x during the public preview of BigQuery history-based optimization
BigQuery history-based optimization is more than just a static collection of four new improvements; they are a platform for continuous investment in BigQuery’s optimization capabilities
You have the ability to examine which BigQuery history-based optimization was used (if any) in INFORMATION_SCHEMA and comprehend how they affected your jobs
In certain situations, BigQuery may determine that it could be more effective to switch the two sides of a join, lowering the resources used to do that join operation
BigQuery may decide to perform the selective join earlier if it comes after less selective operations like aggregations or other joins