DevOps in the Era of Hyper-Convergence

IT organizations have been told they need to tightly integrate application development and IT operations to succeed. But achieving that DevOps goal can be difficult for many companies because their IT operations are fractured. Indeed, the very history of IT infrastructure is defined by the separate management disciplines that grew up around silos of compute, storage and networking resources.

But a new generation of hyper-converged systems is shifting IT operations to be able to respond to application development requirements on demand. In fact, a study from Forrester Consulting, “A Practical Guide for Evaluating the Value of VCE Converged Infrastructure for Data Center Modernization,” finds that improved management and flexibility is the most important reason cited by IT organizations for making the shift to our hyperconverged platform.

The biggest frustration developers have with internal IT organizations is the amount of time it often takes to allocate resources. It may take only a few minutes to provision a virtual machine, but provisioning the associated storage and networking resources still can take weeks. For that reason, many developers often try to bypass the IT operations team altogether.

The key to making IT operations more responsive to the needs of developers is to break down the IT infrastructure silos that now exist. Instead of isolated teams of compute, storage and networking specialists, a hyper-converged platform integrates those functions under a common control plane. IT organizations can opt to either enable those IT infrastructure specialists to collaborate more easily with one another, or they can opt to consolidate those functions under an IT generalist.

Regardless of the approach taken, the IT organization becomes more agile by several orders of magnitude. That means not only can it adapt more rapidly to the changing needs of the business, but any pretext a developer might have had for using a rogue external cloud service has been eliminated.

The IT operations team is still responsible for putting the guard rails in place that make sure IT resources are consumed efficiently. But the days when every IT knob must be set manually before a developer could begin writing code are coming to an end. Hyper-converged systems allow the IT operations teams to define polices that are then implemented automatically. Any necessary changes can be substantiated across any number of nodes in a matter of minutes without requiring the IT operation team to learn how to write code or, conversely, force developers to waste their time writing a custom script to automate the provisioning of IT infrastructure using arcane application programming interfaces.

Much like DevOps, making the shift to hyper-converged platforms is as much about changing the IT culture as it is the technology. A well-designed hyper-converged architecture enables an IT organization to make that cultural shift at its own pace and speed. And while that hyper-converged platform may dramatically reduce IT operational costs, there are plenty of other ways for IT personnel to add more value to the business. But, like most things, there will always be a period of adjustment before everyone appreciates just how much better their daily job actually can be.

About the Author: Todd Pavone