Many approaches to processing and storing petabyte-sized data are in use today, with the focus suitably at the 'Top' of each technology 'stack'. Popular examples include Hadoop, Lustre, and OpenStack. What is often overlooked is the need for a trustworthy layer at the *base* to provide a reliable and scalable foundation for the top layers.
In this talk we will cover the base-layer options available for the most common massive-scale storage solutions, and discuss real-world results of production deployments.
