Scaling Crashlytics: Designing Web Backends for 100 Million Events per Day

Working on the next hottest startup? Building an app that you know will take off?

Your backend web architecture is one of the most critical factors in your long-term success. It dictates your success in launching the next big customer, the amount of work your team will have to do in the face of scaling, and even your ability to bring new features online. Decisions you make now can save literally months of work over the coming years and give you the power to deliver features the competition can only dream of building.

The best part? It’s really easy. We’ll explore fast-message-queueing, a web architecture pattern we’ve employed at Crashlytics since day 1 to process millions of data-points a day. Learn how to design the backends used by some of the most successful Silicon Valley (and Boston!) startups as we walk through code examples and a live demo.

Topics include:

• Job queuing vast amounts of data
• Real-time frontend updates/notifications
• Monitoring performance metrics and troubleshooting
• Scaling from a single server to hundreds using the same pattern
• Lessons learned in production

The talk will be given by Jeff Seibert, the Co-Founder and CEO of Crashlytics, the leading provider of real-time crash detection and analysis for mobile applications. In 2007, Seibert co-founded Palo Alto-based Increo and served as its COO and Architect until its acquisition by Box.net (http://box.net/) in August of 2009. He subsequently oversaw the integration of Increo’s document processing, preview and annotation technologies into Box’s CCM platform.

Over the past 5 years, Seibert has designed, implemented, and scaled cloud-based and distributed processing systems intended to handle 100 jobs per day to 100 Million jobs per day. He holds a Computer Science degree from Stanford University and was selected as a Mayfield Fellow in 2007.