“In a box” has become a marketing phrase used to imply a lot of different elements combined together in a neat, single package. The phrase has been used in many other ways before, often implying a restrictive situation or feeling “boxed in.” Then there is the infamous skit from Saturday Night Live with a more literal meaning. In IT, though, "in a box" has a bit of a history.
https://www.scalecomputing.com/uploads/general-images/box.jpg
Brief IT History Overview
Information technology can be extremely complex, even looking at the basic hardware components on which it runs: the infrastructure. Integrating the most basic compute elements such as CPU and RAM with storage, networking, operating systems, hypervisors, and security elements can often require multiple technical experts and days, if not weeks, of work. The costs can be very high, which is great news for the system integrators and service providers often employed to assist with these projects. But does it have to be this way?
No. And it wasn’t always this way. Mainframes used to rule the IT roost before the rise of the standalone server. The mainframe was the massive, powerful machine that could handle all the computing needs for an organization. It wasn’t perfect, of course, and it ultimately gave way to the greater flexibility of the server. The limits of the server then lead to storage area networks, clustering, and more advanced concepts like grid computing.
The server model eventually required virtualization to alleviate the overburdening costs of maintaining physical hardware for each server. Virtualization helped, but it was only the beginning. The complexity of combining all of the various components was still a huge cost sink for organizations. The cloud looked promising but was still very expensive. The only real answer seemed to be the “in a box” movement, otherwise known as converged infrastructure.
The Variations of “Converged”
The concept of converged or “in a box” infrastructure has been around for some time without really catching on, mainly because it failed to deliver on promised value. Here are some of the variations:
The Pre-Configured Rack
Often it is all of the usual components (mainly servers and storage) plugged into a rack and pre-integrated with updated drivers, operating systems, and hypervisors. It’s really no different than what you might put together yourself. Someone has simply put it together for you and priced it as a package. It saves you the trouble of having to choose the individual components separately and wonder whether they are compatible.
Cloud in a Box
This is built on the pre-configured rack concept but goes beyond the hypervisor in pre-configuring cloud services in addition to the virtualization layer. This is designed to allow organizations to easily implement a private cloud on-prem. Like the pre-configured rack, this is more or less what you would build yourself from various vendor components, just pre-configured for you.
Converged Infrastructure
This is a broadly used term but most often refers to some of the datacenter components being combined into a single appliance. This could be a combination of simple server and software-defined storage (SDS) or perhaps networking as well. What separates these from the pre-configured racks is that they are generally sold as a single vendor appliance rather than a collection of different vendor components. That being said, “converged” solutions are generally designed to be a hardware platform for a virtualization hypervisor from a different vendor.
Hyperconverged Infrastructure
Like converged infrastructure, hyperconverged combines various components into a single appliance but also adds the hypervisor. The hypervisor is not a separate vendor component as in converged infrastructure, but is a native component to the single-vendor solution. Hyperconverged provides the most complete single-vendor appliance delivering out-of-the-box virtualization with single-vendor support. Both the converged and hyperconverged appliance-based solutions usually have the added benefit of being easier to scale out as needed.
HC3 Hyperconverged Infrastructure
The HC3 solution from Scale Computing is a true hyperconverged infrastructure that is often referred to (sometimes by us) as a “datacenter in a box”. I've even talked about it often as a private cloud solution (cloud in a box), satisfying most if not all the requirements of private cloud/hybrid cloud for most organizations. It combines servers, storage, virtualization, and disaster recovery into a single appliance that can be clustered for high availability and easily scaled out. It is the easiest infrastructure solution to deploy and manage, which is why it is consistently rated and awarded as the best solution for the midmarket (where ease-of-use is so highly valued).
Even with as much of the datacenter as we have fit into the HC3 architecture, we haven’t combined every possible component. We still rely on additional network components such as physical network switches and power supply systems that are probably best left separate. It is these additional components that create the complete datacenter component and why we have partnered with other technology vendors like APC by Schneider Electric.
Schneider Electric is offering pre-validated and pre-configured datacenter solutions combining Scale Computing HC3 with APC Smart-UPS. This solution provides both the award-winning ease-of-use of HC3 combined with the award-winning reliability of APC power. You can read more about the partnership between Scale Computing and Schneider Electric in our press release and more about the reference architectures on the Schneider Electric website.
While a complete “datacenter in a box” solution may or may not become a reality in the future, we believe hyperconverged infrastructure like HC3 is the right next step toward the future of IT. We’ll continue to partner with excellent vendors such as Schneider Electric to keep providing the best datacenter solutions on the market.
