It’s easy to forget why we started to examine the deeper channels and architectural workings of the software-defined datacentre in the first place. As much as we focus on explaining how and why data storage, compute processing and network intelligence resources all now exist in a more controllable definable space — just exactly why did we decide to undertake this shift to a more virtualised and abstracted world?
It’s all about the application
Ultimately, of course, it’s all about the application. That is to say, it’s all about serving the applications around which we hinge our business models so that they can operate at an unrestrained capacity and, at lower levels, at their most efficient capacity.
Fundamentally, it comes down to creating an environment where the application can control its resources. In practical terms we should point out that the network engineers oversee the actual throughput of resources, but the application has a potentially new prominence and efficiency in the software-defined world.
A new dynamic agility
The software-defined datacentre provides us with a computing infrastructure layer that facilitates a new dynamic agility. The dynamism here comes about as a result of our ability to programmatically define and form both the layers of the network and its mechanical component parts including switches and areas connected to automation intelligence and management.
An inherently connected elastic dynamism also exists in terms of the software-defined datacentre’s ability to manage connectivity, aspects of security and overall performance management. These are features that we would previously have expected to find existing as embedded hardware components — today they can be provided through software so they can be delivered in a more defined and controlled manner.
But step back again and ask why we have a need to be more defined and controlled. Exactly why is this? Because a more modular, more specialised, more efficient and more programmatically configurable network allows up to create a platform where applications are the first class citizens of the total IT stack. Once again, it’s all about the applications.
Simple economics
It’s a simple question of economics, better performing applications equals better customer service equals better profits.
As we bring the benefits of the software-defined datacentre online, the DevOps network engineer and system administrator (sysadmin) function be of key importance as they control and shape the computing fabric our business models depend upon. In practice this will mean engineering a variety of elastically provisioned and intelligently partitioned resources into a dynamically manageable pool of power with a defined degree of remote programmability for onward management.
There’s a lot of hardware in software-defined
Of course the software-defined datacentre still runs on hardware. Indeed the hardware vendors who have identified this trend are now building hardware appliances that ship with the required amount of pre-installed, pre-certified, pre-configured and pre-tested software appliances on board to fulfill the needs of the software-defined cloud computing landscape currently forming.
What we can surmise from all of this discussion is that we are building a better computing world for our applications if we engineer these technologies correctly. There is still much complexity here, but the signs point to application dividends in the long run.
Adrian Bridgwater
Latest Posts:
Tags: Data Center, Technology