It’s easy to catch the DevOps bug. Digital evangelists everywhere are claiming it can transform an organization’s ability to deliver software at a pace that matches today’s hungry-for-technology marketplace. And the talk of DevOps at conferences and IT forums is alluring. But it’s also easy to get discouraged when you return home to a different—and perhaps overwhelmingly complex—reality on the ground.
Is it really “just DevOps?”
The IT environment for most enterprises consists of numerous global service providers and multiple interacting and overlapping systems in large legacy application landscapes. Therefore, the impulse to think in terms of “just DevOps” oversimplifies the conversation. Implementing DevOps in an enterprise is more than simply deploying standard tools like Chef, Puppet and Jenkins. It is also more than just a culture and process shift. And lessons that can be taken from cloud-first start-ups are few and far between.
Enterprises need visibility into their entire existing IT value chain. And the deeper you look, the more you realize your enterprise applications actually manifest in the form of services. A journey toward DevOps will inevitably lead you to the principles of design-thinking: empathy, ethnography, divergent thinking and abductive reasoning. Now you have way more to address than just the automation of the build and run functions in application testing and production.
DevOps will redefine how you think about failure
The inherent complexity of an enterprise IT environment creates lots of opportunity for failure—a single shutdown can cost millions of dollars. DevOps promises the achievement of stability and speed, but achievement of stability is not the avoidance of failure. Failure is inevitable. Nicholas Taleb introduced the word “antifragile” to the world when he argued that, instead of chasing zero failure, we should build systems that grow stronger with each attack. This means enterprises that implement DevOps and the mechanics that enable continuous delivery must move their target from avoiding failure to gaining resilience.
A journey to DevOps also starts from agile development. Enterprises that have conquered agile development will go on to continuous integration, continuous delivery, scaled continuous delivery and DevOps. The steps along this journey will reveal the intricacies of implementing DevOps in your legacy landscape with your current set of providers and their global delivery models.
Contracting for DevOps
Contracting with DevOps service providers can offer a promising way to access the talent, skills and innovation required to begin the DevOps journey. But since DevOps adoption in enterprise environments will come with the need to modernize existing application landscapes, the relationship needs to support this more holistic kind of change.
Today, in the best of scenarios, enterprises plan their software release cycles either bi-weekly or monthly. Which leaves hot fixes as the only way to correct bugs, a job left to teams that are forced to integrate these fixes with their ongoing development work. Timing such bug corrections alongside delivering software to production gets trickier and trickier. As DevOps speeds up development cycles, developers can correct bugs much more frequently. A shorter gap between event and reaction means less delay in performance and reduced operational risk. Lower operational risk means lower cost of operation for both the enterprise buyer and the service provider. But are DevOps contractual discussions actually taking advantage of this opportunity?
Service providers will need to supplement their offerings with approaches to modernize legacy environments and use their operational expertise to create intellectual property for DevOps adoption and operation. We are seeing an uptick in application modernization programs over the last 18 months, and a large reason is due to the desire to bring agility and speed into implementing new requirements.
ISG helps enterprises evaluate the need for DevOps and create a plan to tame the sometimes-unruly reality of enterprise IT. Listen as I discuss how contracting for DevOps is changing the way enterprises deliver applications in a recent webinar “Putting DevOps to Work: Contracting for and Developing DevOps Globally.” Or contact me directly to discuss further.
About the author
Prashant works with enterprises to shape their operating models for a digital journey and brings 20 years of expertise in all aspects of applications and platforms, from designing transformations through the whole sourcing lifecycle. Prashant’s experience spans a range of industries, including Financial Services, Telecom and Media, Automotive and Utilities, and a range of geographies, including Europe, the Americas and India. Recently, he helped a Fortune 100 automotive giant consolidate its next-generation sourcing for applications, executing digital transformations right up to application management.