When it comes to devops on twitter I've found @fart's stuff to be pretty great. Particularly this: https://medium.com/weird/why-i-am-the-most-important-devops-... - a little sample below:
"This data is encapsulated in what are called “Data Beans” which are stored on the hard drive as zeroes (Optimal bean or “prime bean”) and ones (Flatbean or “bad bean”). These Data Beans are stored magnetically and retrieved as necessary. If a user requests too much data at one time they can suffer from data poisoning.
The computer’s RAM also stores Data Beans in the cloud. No one knows how this is done or why "