Nick Gentile 12/17/2012
Computer Science major, history and informatics minor, graduating 2015.
This blog entry stems from a course assignment in INF 202 Introduction to Data and Databases. My Topic is http://www.computerworld.com/s/article/9234747/YouTube_scales_MySQL_with_Go_c…
Back in November 2006 the relatively new website YouTube.com was purchased by Google for $1.65 billion. It’s been a year since YouTube’s last interface change but early this month Google redid the interface, making it more basic and streamline like many of their other subsidiaries. The simplicity on the user side is reflective on what is going on behind the scenes. YouTube’s engineers have been creating a new set of software called Vitess which will help open-source MySQL databases work more efficiently in large scale production environments. They are using Google’s programing language Go to create Vitess. Simplifying the amount of data that YouTube’s engineers have to process is significant because it will allow them to make YouTube, and other large databases, run more efficiently in the future.
YouTube has to deal with tremendous amounts of data every day, nearly 72 hours of video are uploaded every minute directly onto YouTube’s file system. However, it uses MySQL to store all of the metadata for each of these videos. Soloman, one of YouTube’s engineers stated that they used MySQL because of its reliability, but when you are managing the amount of data they have to deal with it becomes a tricky process. Vitess is being created to run together with their MySQL database to offer additional management and more efficiency. The engineers at YouTube have been using Go to code Vitess because of its quick compilation times and because of how expressive it is. The programmers can slim code down dozens or even hundreds of lines with Go in comparison to using C or C++.
In short Vitess is allowing MySQL code to run more efficiently and with less memory by consolidating thousands of SQL queries into smaller batches, making it so that MySQL can use fewer resources. This is significant because of huge quantities of SQL queries YouTube receives every day. As Vitess gets further into development it will be interesting to see the impact that it has on YouTube’s servers and perhaps the rest of the computing world. Every year more and more data needs to get processed on the internet and innovation such as this is significant to prevent engineers from getting bogged down.