This week I had the pleasure of sitting down with some critical thinkers concerning the future of cloud computing within the Department of Defense. As an IT firm with deep roots in the DoD and Intelligence community we have recently been tapped to help the Department of Defense conduct research and examine the various possibilities and return on investment that could be gained from highly secure cloud computing solutions.
As I sat and listened to finance types backing the latest and greatest cloud technologies emerging out of Seattle and Silicon Valley and then listened to a briefing on efforts to design, test and deploy cloud based services within the Department of Defense I could not help but think what would happen if the various somewhat duplicate efforts racing to deploy their version of secure cloud computing came together to filter out the common elements in their thinking and designs while allowing each entity to peel off to build the truly unique aspects of their cloud solutions needed to meet their mission needs.
In this fickle Government budget season, which according to many estimates will continue to be fickle until the end of the presidential election, it would make sense for at least some agencies to collaborate on cloud computing in order to conserve precious budget dollars. I have been around Federal IT long enough to know that I am only thinking wishfully and that every Federal entity will in fact spend dollars to architect their own version of cloud computing despite the pressures of impending budget cuts. It is also a fact that when too many entities come together to solve a problem or create a solution in the government arena the chances of achieving critical mass towards a tangible solution will almost not meet the time to market needs of any particular entity.
As we launch into our cloud computing solutions research and architecture recommendations for the DoD we plan to keep a few things in mind:
Availability – since our customer is keen on high availability, we plan to assess and define availability. Cloud computing has taken a few hits over the last few weeks due to highly publicized outages of some well known cloud providers. What the recent outages in cloud service provider services clearly point to is a need not just for infrastructure design for maximum availability but also for implementing what we already know as a best practice. We have to engineer to reduce the impact of human error on high demand services and we know now better than ever that cloud solutions are not immune to lack of configuration management and testing.
Security – this is a big one. It might be more important than high availability when we sum things up. Cloud services have inherent security exposure risks that can have rapid horizontal impact to cloud based services. Security also has a direct impact on availability as well.
Performance – rapid time to mission readiness and execution is highly important to our customer. Cloud computing cannot escape the need for robust infrastructure and bandwidth. We are studying bandwidth to the field and what the as-is limitations are to getting mission services and data to where they are needed quickly. Infrastructure performance is also key. We are looking at hardware, new virtualization technologies, storage technologies and believe it or not some degree of distributed application and data distribution in combination with cloud based assets to speed the ability to mission readiness and circumventing bandwidth challenges.
The above elements are a 50,000 foot view to just some of the things we plan to study, test and evaluate. We will also be looking to collaborate with a similar effort just next door to our project with the hopes of leveraging lessons learned and common ideas with the hopes of accelerating the time to market of several cloud computing prototypes.