OpenStack® is the leading open source IaaS platform, powering many of the world’s most notable science and research organisations. Research and science disciplines comprise some of the most prevalent use cases for OpenStack clouds. High-performance and high-throughput computing (HPC, HTC) require massive scaling and cluster networking; storage, compute and networking access to large volumes of data; and workload and infrastructure manageability. OpenStack software supports these needs today and with ongoing development plans to fill gaps. By managing resources as an OpenStack private cloud, researchers are able to work in environments tailored to their requirements. The dynamic, automated nature of software-defined infrastructure cuts away time wasted on the distractions of setup, and enables researchers to maximise the time they spend on research itself. This paper is intended for HPC system architects and research computing managers that are exploring the benefits of cloud and how to bring those benefits to HPC workloads. It can also be used by current OpenStack users to delve into additional capabilities. This is a deep dive into the important functions, considerations, and further reading. Each section includes user examples that describe real-world architecture and operations. A section on research cloud federation covers OpenStack's role in enabling global, secure resource availability for the scientific community.