I found the article about why Hadoop is the future of database interesting. It is very important for a company like Facebook to track a lot of data because of the way that they can use information to make their service better. By knowing what people are doing on the site and tracking data about how people use it, they are able to make changes that can enhance the service and make them more successful. For Facebook, this success means a large number of users and a lot of use by those members.
By knowing things like how people use Facebook, Facebook is able to make these features convenient and accessible. They can get more advertisers by showing them the data of how many views a potential ad would get. They can advertise on other sites and figure out what kinds of sites they will get the most interaction from on their own Facebook ads. All of these things help them to gain popularity which translates to more revenue.
This is why the necessity for software that can store all of this information is so strong. Facebook decided to give Hadoop a shot, and it worked out for them. Hadoop was able to do what all of the older traditional databases could not do. It was the best option for storing the massive amounts of data that Facebook needed to keep track of. The article mentions adding a more advanced SQL query engine that will help Hadoop be even more productive than it already is.