By Andy Flower
Through our ResearchWire program, the Independent Oracle Users Group (IOUG) sponsors surveys to take the temperature of IOUG membership regarding critical enterprise data management topics. We conduct periodic and annual surveys, such as our data security and data growth surveys, that provide visibility into interesting data management trends. This year we also conducted a survey on cloud computing.
The first thing that caught my eye in the cloud survey results is that more than 40 percent of our members are already developing private cloud computing capabilities. To me, this was a surprisingly large number. Virtualization of database servers has been a growing trend, but the 40 percent figure for managing these servers as a private enterprise cloud seemed high. After all, the concept of cloud computing has only been around for a short time.
So what is the difference between cloud and virtualization? To me, virtualization is a technology that enables us to deploy clouds. Being able to create virtual system resources to more fully utilize physical system resources lays the groundwork for cloud computing.
Cloud, on the other hand, allows us to deploy predefined applications or reference architectures in shared virtual environments. Consumers of commercial cloud computing can have a third party deploy applications in a software-as-a-service (SaaS) model, where the third party provides a predefined virtual environment with a predefined application that we as consumers can immediately begin using or customize and integrate into our enterprise application stack. Or, as is the case with Amazon Web Services, we can simply specify a computing environment with the software we would like to use to build our own applications in the cloud.
For private clouds the idea is the same, except that we own and manage the physical environments internally within our organizations. IOUG members are seeing that providing an internal on-demand computing utility may be more cost effective for organizations in which business units or departments historically have spent their own budget dollars for silos of hardware and software and the staff to maintain it all. Why not take a more shared, utility-based approach to providing computing and application services that has cost advantages yet at the same time remains within the control of the enterprise? In private clouds, organizations can define reference architectures or deliver complete applications and use virtualization to more quickly deliver predefined computing environments for internal use. We can deploy platforms as a service to create development and testing environments that are more consistent with our production environments. We can define standard reference architectures for transaction processing database servers, for data warehouses, for middleware, and for application resources.
So what about commercial cloud computing and SaaS? According to our survey results, IOUG members are not exclusively choosing private clouds over public cloud services. Organizations that have an application or computing need that is more readily provided by a commercial public cloud supplier will still leverage those resources. But there are still a great number of organizations with applications and computing needs that are just not comfortable with having critical business functions outside their control. Does this mean every application and all data must continue to reside internally? No. But neither should it all reside on a public commercial cloud. Private clouds do not preclude organizations from going outside to public clouds.
I am not a prescient analyst who can definitively state what percentage of corporate computing will be done in the public cloud, in private clouds, and on dedicated datacenter machines. This is why IOUG does its surveys annually. The reaction to our first cloud survey has been fantastic, and I am eager for IOUG to do this again next year so we can see what interesting trends form in the adoption of enterprise cloud computing.
Photography by Rawpixel, Unsplash