The physical server has always been the combustion engine of hosting, but end users can now run on either dedicated machines or the alternative fuel of virtualized resources.
Customers usually leased physical servers to host the content and applications we consume on the Internet, but virtualization made it unnecessary to lease entire boxes. Some virtual resources are relatively static, or can be highly scalable and flexible--characteristics typically referred to as “cloud.”
To this point, customers have made a decision to purchase one or the other. They lease a physical machine or rent a virtual or cloud infrastructure service. The choice can be driven by preference or compliance considerations, but for the most part technical requirements dictate the choice. The driving force behind hybrid is finding the most hospitable and optimal venue by matching up requirements with infrastructure.
Dedicated servers, for example, are a great fit for many high-capacity workloads. They excel at running i/o-intensive applications like databases or mission-critical applications like e-commerce or transaction-heavy websites. Similarly, many content-heavy websites or mobile applications run perfectly fine on virtual infrastructure.
In hybrid scenarios, the two come together to perform one job. Hybrid allows organizations to optimize their performance and manage costs by using the type of infrastructure that allows them to meet specific requirements of the use case. In a bit of an oversimplification, it is like getting the best of both worlds: the speed, stability and robustness of bare metal with the scalability, efficiency, redundancy and cost advantages of cloud.
For the most part, hybrid scenarios today are not cleanly automated. A typical hybrid setup is in essence a single infrastructure environment made up of unlike parts. The i/o-intensive requirements will run on bare metal and the front-end web application will run on virtual or cloud. Despite being different types of infrastructure, they are fully integrated and “talk to each other” over the same private network.
But if hybrid is the wave of the future, what’s next?
Things are about to change very quickly. Signs point to the diversity of infrastructure types merging into a single experience on the front and back ends. In contrast to much of what we see today, the future of hybrid is highly automated and tightly integrated.
Spinning up and scaling hybrid infrastructure will be done automatically through a few point and clicks or a fully featured API. It will mature to the point where it becomes a single experience and end users control everything from a single pane of glass. There will no longer be two control panels, logins, SLAs and support contacts. The end user will experience one user interface, and provisioning, management and billing will run on the back-end off the same infrastructure delivery platform.
In the early days, customers ran with a single form factor, and infrastructure scaling was a manual process. The emergence of virtualization and cloud made things more complex, and infrastructure became dispersed and disaggregated. Hybrid will take us back to the future with the various pieces consolidating into a single experience – much like that standalone dedicated physical server that got it all started.