MySQL.Conf.AU

The MySQL Miniconf at linux.conf.au is being held on the 15 January 2007.

Schedule

08:00 - 17:30

Featuring The Answer Guys Morgan Tocker and Arjen Lentz, from the MySQL Support Team

08:00 - 09:30

LCA registration desks open

09:30 - 10:30

LCA Keynote

10:30 - 11:00

Morning Tea

11:00 - 11:10

Welcome to mysql.conf.au - Arjen Lentz, Stewart Smith, Colin Charles

11:10 - 11:55

Know your Storage Engines - Arjen Lentz

11:55 - 12:30

10 Do's and Dont's for MySQL Cluster - Jonathon Coombes

12:30 - 14:00

Lunch

14:00 - 14:45

Self-healing databases: managing schema updates in the field - Jonathan Oxer

14:45 - 15:30

Optimizing Stored Procedures - Guy Harrison

15:30 - 16:00

Afternoon Tea

16:00 - 16:45

Refactoring a production database - Morgan Tocker

16:45 - 17:25

MySQL troubleshooting - Laura Thomson

17:25 - 17:30

End of mysql.conf.au

Abstracts

Know your Storage Engines - Arjen Lentz

Different storage engines have different strong points - use them to your advantage! You can choose per table. We'll look at what's possible (with examples), what's optimal, and what would be a bad idea. Includes funky tricks!

10 Do's and Dont's for MySQL Cluster - Jonathon Coombes

Many people are looking at MySQL Cluster as a general high availability solution to their database needs. They quickly decide to drop their existing architecture and implement a new NDB cluster instead. After the first week of trying to get it up and working it seems to be up and ready to populate. Then on import, errors come up and it does not seem to want to take the data? Why is there so many problems in getting this supposed simple database up and going - it was very easy before with MySQL?

This presentation looks at some of the most common mistakes people make in implementing a NDB cluster architecture and why they are making it hard for themselves. What things should be avoided and why does it matter that I have only 256MB of memory, if I can spread it across 24 machines in a large room? It also looks at some of the things that should be done, that is often neglected in setting up a clustered database. Is network speed important? Should I use shared storage instead? Do I go with GigE or SCI (or any other high speed transport method)? These are some of the areas that will be addressed in this presentation.

Self-healing databases: managing schema updates in the field - Jonathan Oxer

Database schema update management is a problem that is overlooked by many web application developers initially, but once you have a number of deployments in the field it rapidly becomes a major headache to propagate schema changes in synch with your PHP, Python, or Perl application code.

Internet Vision Technologies has developed a technique for "self-healing databases" for the SiteBuilder web application suite, allowing updates to occur automatically in the field whenever new versions of the application are pushed out. This technique has been used successfully on a number of large scale deployments such as the Siemens intranet which runs on MySQL and uses over 2500 tables. The talk will cover the overall methodology with specific examples from the SiteBuilder codebase.

Optimizing Stored Procedures - Guy Harrison

This talk will look at the most important factor influencing MySQL5 application performance - the optimization of SQL statements and stored procedures.

We'll first consider the optimization of simple, single table SQL statements by using effective indexing techniques and database design. We'll then consider how to optimize joins and more complex queries involving derived tables, views, ordering, grouping and DML.

The first step towards optimizing MySQL5 stored procedures and functions is to optimize the SQL contained within these programs. However, as with any programming language, the design of loops, conditional (IF and CASE) and other structures can have a significant effect on overall execution time. We 'll look both at how to optimize the stored program code, and how to optimize the use of SQL within that code. Then we'll consider some cases in which you can use a stored programming approach as an alternative to hard-to optimize SQL.

Finally, we'll look at how new features in MySQL 5.1 can be used to further improve the performance of your MySQL server.

Refactoring a production database - Morgan Tocker

Changing the structure of your tables is one of the more painful changes you can make to any application. The release of MySQL 5 brings views, triggers and stored procedures.

Using a combination of these technologies I will show a series of examples of how to migrate your data while maintaining a view of it in the older format. Your application will continue to work with minimal change, and without lengthy migration period.

MySQL troubleshooting - Laura Thomson

In this talk, we'll work through the process of troubleshooting problems on your MySQL server. Topics include:

  • Tracking the problem to your database
  • Working out what is slow and why it's slow
  • Solutions to common problems

Resources

CfP

Miniconfs/MySQL (last edited 2007-01-14 00:33:47 by ArjenLentz)

  • Immutable Page
  • Info
  • Attachments