There are many definitions and connotations for the word ‘vapor’. In relation to IT one might think ‘vapor’ to refer to ‘vaporware’, software that has been promised, but is never delivered. However, in the context of this project, ‘vapor’ refers to the diffuse water present in the atmosphere that condenses to form clouds.
Vapor itself is a technology framework and governance arrangement to facilitate both the sharing and compartmentalization of cloud-related virtualization technology in research and instructional settings. It exists to facilitate the sharing of knowledge, personnel, and technology among its collaborators.
The goal of the Vapor Cloud project is the formation of a geographically distributed infrastructure to facilitate both shared and autonomous clouds. Those clouds will support instruction, research on cloud computing and research using cloud computing. Vapor Cloud will integrate multiple autonomous cloud infrastructures, distributed over several geographic locations on a shared high-performance network fabric. It will leverage common identity management (GTED) together with a standard API (OpenStack) as well as virtualization technologies such as Hyper-V, RHEVM, VMWare and EC2 to provide resilient and flexible computing resources to the Georgia Tech academic community.
Why not a single cloud? By allowing for a diverse implementation of multiple clouds each meeting differing use cases, experimental risks can be compartmentalized and group autonomy (academic units, research groups) can be preserved. But by federating the cloud, sharing of resources amongst groups can be facilitated as well as further collaboration with PACE for access to high performance computing, big data storage, and other research support areas. Further, federation with other institutions and/or cloud service providers will present even more avenues to collaboration.
Vapor is a collaborative effort between the College of Engineering, the College of Sciences, the College of Computing, RNOC and OIT.