Part 1 of "Understanding How MySQL Works by Understanding Metadata", presented by Sheeri K. Cabral (The Pythian Group) and Patrick Galbraith (Lycos Inc.). This was a 3-hour tutorial.
The PDF of the slides can be found at http://technocation.org/files/doc/2009_04_Understanding.pdf.
From the official abstract at http://www.mysqlconf.com/mysql2009/public/schedule/detail/5682:
We have spent countless hours researching over 1,000 pieces of metadata. In the process, we have learned a lot about how MySQL works, and realized that it was a pretty good learning method.
Examples: Understanding the “query_cache%” system variables and “Qcache%” status variables helps us learn about the query cache—what it is, when it is used, how to examine query cache efficiency, how to tune the query cache. This relates to the GLOBAL_VARIABLES and GLOBAL_STATUS system views and corresponding SHOW commands.
The CHECKSUM field of the TABLES system view in the INFORMATION_SCHEMA DATABASE seems straightforward—it holds the checksum. But when is that field updated, and for which storage engines?
Like CHECKSUM, there are many storage-engine features that are hiding in plain sight. By reverse engineering MySQL’s metadata, we will show you many of these features.