Statistics about tables, columns and indexes are a critical component in producing good query plans. This session will examine the differences between random AMP sampling, full statistics collection, and USING SAMPLE collection, and will share details about how each of these 3 approaches works internally. Statistics histograms, their uses and limitations, will be explored. You'll get some insights into collecting statistics on multiple columns, on indexes, on and on the system-derived column PARTITION. Statistics extrapolation will be touched on briefly, and the session will wrap up with some general recommendations for collection statistics.
- Learn the differences between the collection approaches
- Understand how statistics extrapolation functions
- Get useful tips about statistics
Presenter: Carrie Ballinger, Sr. Technical Consultant - Teradata Corporation
Audience: Database Administrator, Data Warehouse Architect