Building a private cloud for an academic environment

While the perceived and real technical and soio-technical concerns about public cloud solutions may be discouraging many organizations from migrating and/or using public clouds, many organizations are resorting to solutions involving private and community clouds. The increasing trend of private clouds is also making inroads into academic circles where a large amount of data may have very specific privacy and security constraints. However, a large number of academic environments, especially individuals and small teams of researchers, are unlikely to have the required knowledge, skills, and time to fully understand and appropriately address the selection of appropriate hardware and software, and configuration and deployment challenges. In order to provide such a guideline, we decided to carry out a project aimed at building a private cloud using open source solutions like Eucalyptus and OpenNebula. The key objective of this project was to develop a reference guide for designing and implementing a private cloud using open source software, and the focus was primarily on practical aspects concerning design, installation and implementation. The project was carried out by Zoran Pantic under my supervision. Zoran, being an IT professional, leveraged his huge amount of IT infrastructure knowledge and skills to build a private cloud that we plan to use and extend for our research and education on cloud computing. Zoran and I plan to further refine and publish the guidelines from the project report he has written. If you are interested in getting the report in its current form, please write to us.
This effort has concluded that private clouds tend to be more customizable but at a greater cost in terms of efforts, time, knowledge and skills required. Such effort requires steep learning cover and involves several challenges when evaluating, installing, managing and maintaining a private cloud. While public clouds are mature and well supported, there are several privacy, security, and availability concerns which need to be addressed. One solution for those interested in building, using, and maintaining a private cloud in an academic environment is to start with a public cloud with data that is not highly secure and private and learn about the available technologies for building a private cloud by experimenting with them. And the technologies are considered mature enough and required knowledge and skills are available, then build a private cloud and move from the public to private clouds.

This entry was posted in Cloud Computing, Container Technologies, Cyber Security, Data Exfiltration, Private Cloud, Security, Software Architecture. Bookmark the permalink.

Leave a comment