Although DevOps is still fairly new in the enterprise space, it has already proven to be a powerful solution for improving performance, agility, and productivity for teams and organizations. However, it is still firmly entrenched in the inflated expectations phase of the hype cycle. Big business is simultaneously leery of making a commitment to DevOps, and expecting more from the transition than is feasible over the short term. What’s the reality an organization can really expect as they make the change to a DevOps way of doing IT?
DevOps is no overnight fix
As a DevOps consulting firm to the Fortune 100 crowd, Newt Global aims to make DevOps out of the box a reality for large organizations. To make that happen, they have to know what the phases of implementation really look like—and have the ability to manage client expectations. Venkatnadhan Thirunalai, DevOps Practice Leader at Newt Global, spoke about the typical journey at DOES 2016 in San Francisco. He said it is not unusual for enterprises to get a little ahead of themselves.
“They want to do everything fast. They implement a lot of automation. Very soon, they realize that automation is not good enough. They need to bring in engineering practices. They want to Dockerize, to orchestrate the builds. They want to do data visualization.” The point when the client realizes that DevOps is more than automation is when they graduate from infancy and enter what Venkatnadhan calls, Zone Two. That’s when they may see gains but also face another round of disillusionment.
Faster and better is not enough
“When they scale up the applications, they see, ‘Hey, we are doing stuff fast and having the practices, but we are not getting the business benefit.’ Then they graduate to Zone Three.” This is where DevOps becomes more refined and several fundamental shifts may occur. The easy wins of better tooling and even cultural shifts and engineering practices can’t take the client over the finish line. They must be ready to dig in and make some more fundamental changes to their application portfolio and infrastructure.
“They decide to rearchitect the applications for the cloud and microservices, container orchestration, and so on.” Such changes can entail making a substantial investment of time, money, training, and other resources to allow an organization to fully leverage DevOps. But it’s the phase when it all begins to pay off. In Zone Three, “That’s when they start getting the real agility and benefits of DevOps.”
Phasing in DevOps should start with the long term outcome in mind
Large scale organizations are already using a roadmap to understand the importance of a phased transition for the cloud. Now, Newt Global and similar consulting firms are attempting to provide the same clarity and guidance for enterprise DevOps. A more considered approach may well be the key to preventing the backlash that seems inevitable from big businesses that try and seem to fail at DevOps. Perhaps, in many cases, they didn’t fail—they just gave up too soon.
Why it shouldn't be that hard to deliver DevOps at scale
Continuous integration and continuous delivery are the cornerstones of DevOps transitions
When it comes to implementing ALM, culture is the key