Hash Joins and Bloom Filters

Early filtering of rows, to prevent their expensive flow through the various stages of an execution plan, has always been crucial. In this presentation we will explain the concept of Bloom filters. These filters can be used by the Cost Based Optimizer in hash-joins to ‘early filter’ rows. We will start of by explaining how a hash-join works, both serial as well as parallel. And then demonstrate how a Bloom filter can speedup this type of join by discussing a few example queries.

Wednesday, 2016-11-16
1:00 pm - 1:45 pm
Seoul
beginners
English
Database
Koppelaars

Toon Koppelaars

Oracle Nederland BV
Oracle OAK-Table
Toon has been part of the Oracle ecosystem since 1987. He is currently a member of Oracle's Real World Performance team. The RWP-team troubleshoots application performance issues in and around the DBMS. The way applications currently use (or rather, abuse) the DBMS, is often at the root of these performance issues. Prior to joining the RWP team, Toon has been mainly involved in database application development. His special interests are: architecting applications for performance and scalability, database design, and business rules / constraints modelling. He is a long-time champion of the Thick Database paradigm.

Further lectures of Toon Koppelaars

The NoPL/SQL and Thick Database Paradigms, Part 2
Thursday , 17.11.2016
St. Petersburg
4:00 pm - 4:45 pm

These lectures could also interest you