MySQL Query Anti-Patterns that can be moved to Sphinx

Trends in Architecture and Design
23 April 2:20pm - 3:10pm @ Ballroom A

There are some types of queries can become a bottleneck in your MySQL instance as your dataset and traffic increases. Not only full text and wildcard search, but also full table and large dataset scans and sorts. In this talk I will share hands-on experience with identifying these and moving them to an appropriate tier in the stack such as Sphinx. Advanced techniques such as sharded indexes for parallelism, and in memory scans will be discussed in detail.


Vladimir Fedorkov
Senior Engineer, Blackbird, (formerly PalominoDB)
Vladimir has been helping people to scale their LAMP applications for more than 10 years, with a focus on performance optimization, architectural choices and deep knowledge of search and MySQL technologies. As a team lead on the PalominoDB European team, he provides high-profile remote DBA operational support and performance tuning on large, distributed open source database platforms. In the past he was worked with high-performance systems as a developer and team leader. He's spoken in the past at numerous OSS conferences and meet-ups in the US, EU and Russia. In his off-duty time he enjoys traveling, bicycling and photography.