The Dimensions of Cloud Computing

From MIKE2.0 Methodology

There are innumerable routes into leveraging the internet in the enterprise under the banner of cloud computing. Yet many get locked into a single slice of the multi-dimensional possibilities and define cloud computing that way. In this article, I will present the three dimensions of cloud computing possibilities to help ensure the foray into the cloud is well informed.

First, it’s worth noting cloud computing is perhaps at the peak of hype which could be skewing its prominence in our lives. However, this may be some chicken/egg dynamic since CIOs, under cost pressures, have proclaimed a desire to have upwards of 50% of corporate databases in the “cloud” in the next 5 years. The cloud represents economies for corporate systems. It represents the inevitable. Cloud computing has the potential to change information technology more than any other single discipline ever has.

Cloud computing brings the benefits of speed to provisioning, reduced in-house staff and only paying for what you need. However, it also brings a defined set of challenges starting, of course, with security. The inevitable integration with the inevitable on-premises systems can also be a barrier, especially the performance of the integration. Smart deployments take this into account and co-locate systems with a high degree of sharing needed. While reliability statistics can arguably match or beat what an on-premises solution will bear, it can be unnerving to some to have an outage recovery be completely out of their control. Others see outages as something that happens to all systems and they don’t want to staff the skillsets to perform recoveries.

Most studies of cloud computing will tell you the above, but these seem to refer only to public cloud deployments, which is only one choice in the first important dimension of cloud computing which is public/private. Public cloud computing is the hosting of systems at an external provider who provides such systems as part of their business model.

My favorite definition of cloud computing comes from The National Institute for Standards and Technology (NIST), who defines the 5 cloud computing pillars as On-Demand and Self-Service, Broad Network Access, Resource Pooling, Rapid Elasticity and Measured Service. As long as these are present, it is cloud infrastructure. You can achieve this with a private cloud as well – a “virtualized” arms-length arrangement with a technology wing of your company. Many of these cloud deployments are still on the path to the 5 pillars and the use of “cloud” can be rather euphemistic. It should be noted that a third option is pooling private clouds to provide a hybrid cloud infrastructure.

Public clouds provide a much wider range of infrastructure possibilities, given it is their business model to serve a variety of client needs. Compared to private clouds, public clouds have simpler boarding processes. They have short-term cost benefits and provide obvious rationale to make all costs operational. Some companies, however, are sacrificing the short-term benefits and establishing private clouds now.

The second dimension of cloud computing has to do with what you are getting out of your cloud infrastructure. It could be pure, unsoiled infrastructure and that is IaaS – infrastructure-as-a-service. IaaS gives you access to servers, storage and networking over the internet. Amazon famously provides this level of service. Adding the middleware and operating system to the servers, storage and networking is PaaS – platform-as-a-service. Azure, force.com and others fall into this camp. Continuing up the services chain, we get to the fullest package – SaaS – software-as-a-service. With SaaS, applications are additionally provided. With SaaS, you just “use.”

As individuals, many of us consume SaaS applications on the cloud daily such as Google Apps, DropBox, and iCloud. Perhaps the most prominent enterprise package using a SaaS model is salesforce.com.

The third dimension of cloud computing is also addressing “what” you are putting on the cloud and this is where the possibilities have become numerous. Take a data warehouse, with its sourcing/integration/transformation, data storage and data access layers, broadly speaking. Any one, two or three of these can be placed into the cloud, irrespective of the others. Many are starting with the database itself and working “out” to the data access layer.

The voyage into the cloud can begin from any of the combinations of the dimensions: public/private, IaaS/PaaS/IaaS and the chosen layer. It should be a mindful evolution to multiple forms, with lessons learned and absorbed along the way that the organization can take through the worthwhile journey into the cloud.