Star/join query optimization

摘要: Unwieldy star/join queries are performed more efficiently using a filtered fact table. Suitable include with large table joined multiple subsidiary dimension tables, where indices exist over join columns. The query is analyzed to prepare plan for the accesses. This supplemented by adding nested loop operations, inner and outer an index scan of column also filtering records resulting from joins sequence dynamic bit vectors, ultimately yielding list probable records. further fetching these construct distilled which used, instead original table, execute in considerably less time. If desired, other competing approaches may studied provide cost estimates, least costly approach being actually implemented.