Presentation Title: Bloom Filters
Speaker: Mr Christian Antognini Oracle ACE Director OakTable
Presentation abstract: A bloom filter is a data structure used to support membership queries. Simply put, a bloom filter is used to test whether an element is a member of a given set or not. Since Oracle Database 10g Release 2 bloom filters are used in various situations. Unfortunately, not much information about their usage is available in Oracle Database documentation. The aim of this presentation is to explain not only what bloom filters are, but also, and foremost, to describe how the database engine makes use of them. Specifically, it explains how the database engine uses bloom filters to improve the performance of joins (including the particular cases with Exadata and In-Memory Column Store) and to implement join-filter pruning.
Presentation begins: 07/12/2016 14:10
Presentation duration: 50
Presentation content level: 2 (1 = indepth, 5 = strategic overview)
Audience experience: Intermediate
Audience function: Product Expert
Speaker biography:  Since 1995, Christian Antognini has focused on understanding how the Oracle database engine works. His main interests include logical and physical database design, the integration of databases with Java applications, the query optimizer and basically everything else related to application performance management. He is currently working as a senior principal consultant and trainer at Trivadis in Z├╝rich, Switzerland. If Christian is not helping one of his customers get the most out of Oracle, he is somewhere lecturing on application performance management or new Oracle Database features for developers. In addition to classes and seminars organized by Trivadis, he regularly presents at conferences and user-group meetings. He is a proud member of the OakTable Network and an Oracle ACE Director. Christian is the author of Troubleshooting Oracle Performance (Apress, 2008/2014) and the co-author of Der Oracle DBA (Hanser, 2011).
Related topics: Database: Performance & Internals
Hall: Hall 5

BloomFilters_UKOUG_2016.pdf1.2 MB23/11/2016 16:57:22