My name is Zak Greant. I work for MySQL AB as their community advocate - I am also helping to organize the content for our upcoming MySQL User Conference (http://mysql.com/events/uc2004) that is happening in April.
Our deadline for choosing the final list of sessions is a couple of days away. We have so much excellent content that has been proposed that we need some help from the community in choosing the content that is best for the event.
Here are seven advanced/internals talks that we are considering. Please let me know which talks you would best like to see.
Notes:
* We will be posting addition information on the accepted sessions to http://mysql.com/events/uc2004/sessions.html within the next day.
* Even if you are not going, your feedback is valuable to you as we will work to post papers from most of the accepted sessions.
* We are open to suggestions for other content in the form of BoF sessions and lightning talks (see http://www.mysql.com/events/uc2004/speakers.html)
The session titles are:
* Connecting Data to MySQL, writing a simple Storage Engine
* Hack Proofing MySQL
* MySQL performance comparison on RedHat linux, Debian GNU/Linux and FreeBSD
* Performance tuning for Innodb storage engine
* Protecting Private Information with MySQL
* Speed, Flexibility, Security: The MySQL solution to a large-store XML dilemma
* The Badger Project: Improving MySQL Performance on Modern Computer Architectures
Connecting Data to MySQL, writing a simple Storage Engine
Noted community hacker Brian Aker demonstrates how to write your own custom storage engines for managing specialized types of data.
Hack Proofing MySQL
MySQL databases are a common point of attack because they typically hold the most valuable information for an organization. Because of this, locking down your MySQL databases is critical. This presentation will cover how a hacker approaches your database, the most common attacks, the most successful attacks, and how to prevent yourself from ending up as a victim. Secure your MySQL server involves properly configuring the database, educating your developers, and developing your applications to access the database securely. We will cover attacks such as SQL Injection, password brute-forcing, buffer overflow, and privilege escalation. Attendees will learn: 1) How attackers attack a MySQL server. 2) How to securely install and configure MySQL. 3) How to securely write applications for MySQL.
MySQL performance comparison on RedHat linux, Debian GNU/Linux and FreeBSD
When talking about databases, we normally do not think about the underlying environment on which the application is running, but this can represent a great advantage or drawback if we consider that each programmer, the application developer and the operating system developer, have to focus on his own work and depends on the work of the other.The two components have to be fine tuned to smoothly work together. When WillyStudios.com had to decide which one was the best platform to deploy its work on, we tested each environment thoroughly. In this talk we are going to present the results showed by our survey and discuss which ones are the best methods to have higher performances from the same database environment on different platforms.
InnoDB Performance Tuning
This session will demonstrate techniques for optimizing the performance of the InnoDB storage engine.
The speaker will cover both MySQL server configuration as well as best practices for application development and OS tuning/hardware tuning.
The advice provided will be supported by benchmark results.
Protecting Private Information with MySQL
Website managers must balance the need to provide customization with the responsibility to protect the private and sensitive information of customers. This talk will explore how some of the built-in encryption features in MySQL can create safe, strong and secure systems to guard against hackers and malicious insiders. It will begin with some simple examples and culminate with an extended example describing a website that offers all of the features of a highly-customized website like Amazon.com without keeping any personal information around. It will be the equivalent of having your cake and eating it too.
Speed, Flexibility, Security: The MySQL solution to a large-store XML dilemma
While MySQL databases have been used for years by the LANL Digital Library Development Team, it has been in the creation of our most ambitious project that we have truly recognized some of the unique strengths and features of MySQL. The project -- development of a comprehensive database of scientific journal articles and citation information -- involves converting bibliographic metadata from 8 distinct data suppliers into a common XML format, enhancing the data with links between each of the 50 million articles and their 507 million individual references, and providing search and browse access to the data. In addition to the quantity of data to be managed, other challenges include maintaining flexibility, response time, reliability, fault tolerance, cost, and security. In general, XML data is nested and not normalized, while relational data is tabular, flat and normalized. However, the realization that our XML data is tabular and normalized allowed us to use a relational database for our solution. MySQL was chosen for the following reasons:
1. Cost: MySQL is Open Source.
2. Speed: MySQL has proven to be fast, vital when handling links among 1,435,000,000 rows of interlinked data in 10 virtual tables.
3. Operating system flexibility: Sun Solaris offers the possibility of addressing up to 32 Gigabytes of memory while running the MySQL server at 64 bits, important for caching.
4. Data storage capabilities: We currently manage ~350GB of data; to limit individual table size, we take advantage of merged tables.
5. Fault tolerance: We use MySQL in production and as disk-based backup for our data.
6. Security: MySQL replication is used to protect and update our data; the master MySQL server runs behind a firewall, while slave servers access the data from outside the firewall, in read-only mode. This presentation will give a brief description of the system architecture and will focus on the rationale for choosing MySQL and the optimizations applied to system components.
The Badger Project: Improving MySQL Performance on Modern Computer Architectures
This talk will present the research results obtained in the context of the Badger project, a collaboration between MySQL and University of Copenhagen started in 2003. The goal of the project is to evaluate, and when possible, improve the performance of MySQL so that it leverages the potential of the underlying OS and hardware. The results obtained so far concern (i) the utilization of Linux asynchronous IO, (ii) load balancing on clusters of cloned MySQL servers, (iii) instrumentation of MySQL to monitor performance, and (iv) MySQL performance on Itanium-2.
-- Zak Greant MySQL AB Community Advocate
-- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
