MySQL Conference 2010 – Real-World Problems
The MySQL Users Conference is the ideal venue for bringing together MySQL creators and developers, DBAs, programmers, curious minds, tinkerers and “tinkettes”… Basically, the entire “geek crowd” united in a big explosion of knowledge surrounding the MySQL ecosystem. In fact, as I’ve been saying for a while, the slogan “MySQL is an ecosystem, not a product” is going to catch on… Actually, it already has. Everyone is echoing this sentiment.
The conference is always open for companies (users, partners, or developers) to present case studies or real-world problems. This is a particularly interesting part of the event, as it shows how other users are facing and overcoming obstacles. Besides sharing knowledge, it brings different perspectives and creative solutions to the community. After all, it’s very dull to live in an environment where only the vendor has the ultimate word and absolute authority. Knowledge sharing and collaboration are key features of the MySQL community.
Ning (ning.com)
Ning is a social network platform that connects people with shared interests. Using simple and practical tools, users can create a portal with a blog, forum, member management, photos, videos, and more. For instance, check out our friend Sarah Sproehnle’s network at EveryThingMySQL.ning.com—a great example of how Ning works and the amount of information it needs to manage.
Talk: “Faster than ALTER – Less Downtime”
The talk titled “Faster than ALTER – Less Downtime” suggests that, based on Ning’s experience, it is much faster and more efficient to export data and then re-import it using LOAD DATA INFILE, rather than relying on ALTER TABLE.
Chris Schneider (Ning) explained that they deal with MyISAM tables over 70GB in size. In such cases, ALTER TABLE becomes unfeasible. Their adopted solution: export the data, apply the necessary schema changes, and then use LOAD DATA INFILE.
Commentary:
It’s important to remember that MySQL/MariaDB processes schema changes by creating a temporary copy of the table. In other words, a second copy of the data will exist either way. However, the way MySQL/MariaDB handles this process is slow and complex. For operations like index creation or column changes, Ning’s method is indeed preferable.
Of course, this is an extreme measure, justifiable mainly for large tables (say, over 2GB to 5GB). But it also doubles as preventive maintenance: depending on the storage engine (InnoDB, MyISAM, Maria, PBXT), it can reduce fragmentation, reorganize indexes, and remove deleted rows. Two birds with one stone—or rather, one well-aimed strike.
Visit our Blog
Learn more about databases
Learn about monitoring with advanced tools
Have questions about our services? Visit our FAQ
Want to see how we’ve helped other companies? Check out what our clients say in these testimonials!
Discover the History of HTI Tecnologia
