This article discusses building an inexpensive petabyte database using MongoDB and Amazon Web Services (AWS). The authors chose to use spinning disk instances, which are cheaper than ephemeral storage options. They used AWS CloudWatch to monitor the cluster's performance and found that CPU was the bottleneck, not disk I/O. The authors also tested various query operations on a large dataset and found that they could achieve high query rates using parallel execution across the entire cluster. The article concludes by highlighting the potential for further scaling and optimization of MongoDB and invites readers to share their own experiences with scaling MongoDB implementations.