Answers to the Scaling Challenge: A Case Study With Answers.com on Scaling With Memcached and MySQL

The secret’s out (though of course it’s no secret at all to MySQL users): Memcached will do wonders for any website that must quickly serve up dynamic content to a rapidly growing number of users. Facebook, Twitter, Amazon and YouTube, among the busiest sites today, all rely on Memcached to help them scale out. And it’s well known that Memcached handles 120 million queries per second for Facebook.

These sites invest a lot of manpower into managing their ever increasing number of Memcached servers (some top 50 sites alone dedicates as much as 1/3 of its servers to running Memcached). While it is a fast in-memory caching system, Memcached itself does not scale easily – it over-utilizes expensive memory resources; it does not have an automatic fail over mechanism; it requires extra coding for replication and management. This is where the manpower comes in – to constantly monitor, optimize, and improve Memcached for scalability and high availability.

Not all web businesses, however, can afford (or want) the do-it-yourself route when it comes to Memcached. Answers.com, one of the web’s top 10 growth sites, did not want to be in the business of managing Memcached server sprawl or custom code maintenance. The site aggregates information on millions of topics from over 250 licensed sources, including Houghton Mifflin, Barron’s and Encyclopedia Britannica, and is constantly adding new community‐generated social content. There was only one option for Answers.com: Gear6 Web Cache, a Memcached distribution from Gear6 that has built-in replication, clustering, optimized memory utilization, and advanced management to take the guesswork out of deploying and maintaining a distributed caching tier.

In this session, Dan Marriott of Answers.com and Joaquin Ruiz of Gear6 will discuss the pros and cons of implementing Memcached on your own or going with a commercial solution. They’ll touch on how to:

• Integrate with a high-performance MySQL infrastructure;

• Reduce costs involving servers, rack space, power, cooling, and the time and expertise required to deploy and manage servers;

• Implement authentication and security features; and

• Add capabilities for replication, clustering, optimized memory utilization and advance Memcached management.

• Sponsored by Gear6