Forgive federal CIOs if they feel a little trapped.
Agencies face enormous budget challenges and increased citizen service expectations, yet they have to execute their IT game plans within software development models built for the 1980s. Traditional approaches were created with the best of intentions -- incorporating the insight of a number of disparate internal stakeholders (developers, testing groups, change management committees) to provide checks and balances during system development. It looks great on paper, but in reality, these stakeholders end up batting risk back and forth like a volleyball, and the collective group becomes less concerned with outcomes. The result? Time is lost, taxpayer dollars are wasted, and systems that commercial organizations can roll out in months take years for agencies to deliver.
In the current "Cloud First" federal IT landscape, a DevOps operating model has become a necessity.
[Big deal? Is 1% Improvement Boring, Or A Breakthrough?]
DevOps removes the barriers between development and delivery, cutting across organizational silos and bringing together stakeholders with the common goal of delivering value as quickly and efficiently as possible. It's a set of disciplines that sets out to resolve the friction that builds between internal teams -- most notably the developers and the infrastructure workforce. Deploying a DevOps model isn't a technology challenge as much as a cultural one. Nothing elicits more fear and trepidation within a federal agency IT team than when changes are deployed to the production environment. A bad production release can cost an agency millions of dollars, negatively impact the agency's mission and reputation, and make life painful for the entire workforce.
To avoid the pain, production deployments are often prolonged by negotiations involving numerous supporting organizations. To break through the bottleneck, federal CIOs first need to understand the cultural challenges they're up against, which typically fall into three categories:
No way, Jose. This culture sees the production environment as a complex beast that can't be touched. Saddled with outdated systems, deployments are feared, and changes to production are closely controlled through processes, lifecycle phases, gate reviews, and handoffs. Testing and configuration are manual, time-consuming processes that are done independently, and risk is maintained by the rules of operational prudence, which are ultimately preserved by the operations group. This is the "culture of no" -- prohibiting change where it's most needed.
So you're telling me there's a chance ... The second type of federal agency IT culture sees the production environment as a problem that can be transformed through careful planning, analysis, design, and release management. The developers will eventually get around to training the operations group, but everyone will believe it when they see it. The new concept of operations will be bundled and deployed as one big production deployment with an end-to-end process, but the operations group won't be involved during development. In other words, the will is there, but the collaboration is not.
Let's get it started. Now this feels better. The willing culture sees the traditional production environment as something that needs to change urgently, through calculated risks. The mold of the status quo needs to be broken in order to get capabilities to users in production fast. The opportunity to deploy quickly outweighs the fear of doing so, and the developers and infrastructure teams sing Kumbaya as a single integrated unit focused on outcomes.
Regardless of where an agency's culture sits on the willingness continuum, federal leaders should urge skeptics to think of DevOps not as a management promise, but as a technical and organizational response to their biggest delivery challenges. By combining the disciplines of software development with infrastructure, new forms of automation enable faster delivery of value. Applications and infrastructure are tested from the inside-out as systems are built, improving both reliability and security. Combining portable software architectures with commoditized platforms greatly reduces costs, and merging tools for measuring development and operations means better ways to scale are discovered. DevOps is about combining forces to build confidence, and recognizing the circumstance to change the culture.
The good news is that a number of federal organizations are already jumping aboard the DevOps train. Just this year, DISA and USCIS have embraced DevOps to become more agile and more efficient, and agencies like the VA are exploring the model as well. But perhaps the best example comes from NASA, which recently used a DevOps approach to migrate more than one million pieces of content and more than 100 websites and applications to the cloud.
Over time, migrations began taking place more quickly through automated deployments, and both performance and scalability began to improve. Employees can now update content in minutes, compared to up to an hour, and DevOps has been the catalyst for dramatic success, helping NASA break the mold of old delivery processes in favor of better ones.
DevOps has the potential to redefine how government agencies operate by helping them to become more agile and efficient. While there are cultural challenges to overcome, and every agency is in a unique position, this proven commercial model is just waiting to deliver value to both agencies and the citizens they serve.
If the world wasn't changing, we might continue to view IT purely as a service organization, and ITSM might be the most important focus for IT leaders. But it's not, it isn't, and it won't be -- at least not in its present form. Get the Research: Beyond IT Service Management report today. (Free registration required.)