ApacheCon NA 2015

Everybody's getting a private cloud install these days. Look at those racks, all those shiny new boxes, all that juicy compute and storage, just sitting there, running daytime HR or web serving workloads. Wouldn't you love to get your hands on that spare capacity and put it to real use? Burst your Hadoop workloads into it when no one else needed it, without angering the neighbors? At HP Helion we're working on an OpenStack system to add more knobs and dials to your private cloud workloads, and make it easy to figure out how much spare capacity you have, and then put that compute and storage to work when it isn't in use. In this talk we'll outline the problem space, talk about some approaches to spanning existing Hadoop installs and private clouds, and then show the solution we're building out that lets you consume capacity as it's available, and give it back when it's needed elsewhere.