London startup uberVU on storing 5TB of data in MongoDB on Amazon EC2
An interview with uberVU Founder and MongoDB user Vladimir Oane.
What is uberVU?
UberVU allows you to stay on top of the social media game by giving you everything you need in one simple box. From in-depth analytics to detailed research driven PDF reports to show your boss and impress your clients. We do this through an easy to use, WYSIWYG, off the shelf social media management service that empowers small companies and agencies all over the world to be on top of their game when it comes to social media.
Why did you choose MongoDB?
We started to look at MongoDB when we began to think about the possibility of filtering data. We collect a lot of social media mentions for our customers and the ability to slice and dice it based on sentiment, language or social network is key. We needed to answer questions like: “What are the negative mentions on Twitter in English from Australia from people with more than 500 friends … from 2 months ago?” That’s a complex, real-time query to handle and you need a lot of hacking to get it to work with NoSQL databases. But Mongo seemed to handle it quite easily and we slowly moved all of our data to Mongo.
How have you scaled your application?
We are using a lot of EC2 instances. We are aggregating and filtering around 20M+ items a day that we store in multiple Mongo databases. Since we started using Mongo a few months ago we got to 5T of data. We are operating on more than 50 Amazon instances running MongoDB.
How has using MongoDB allowed you to evolve your application and your data?
MongoDB allowed us to scale a lot easier than in the past and to use a SQL-like approach to data retrieval while maintaining the speed of a NoSQL database.
If you want to learn more, next week uberVU will be giving a live webinar about their MongoDB implementation. You can register here.