As we head into the second half of 2017, it is safe to say that the adoption of DevOps is mainstream. Organizations increasingly rely on modern software development and delivery as part of their business model. Interest and adoption of DevOps is happening on a much broader scale, especially within the enterprise.
Organizations across industries are facing dynamic and highly competitive marketplaces driven by disruptive companies. These companies – the Unicorns - have entered markets with new business models powered by technology and software that can create or destroy billion dollar businesses. One of the prime examples is Apple and iTunes, which basically drove the CD industry into non-existence.
Disruption is rampant and the S&P 500 charts indicate that the average company’s tenure is shrinking. By 2030, it’s estimated that 75 percent of the companies on the S&P list will be new companies . It is more important than ever for enterprises to understand the changing landscape and adapt quickly.
As future-thinking businesses are trending towards DevOps adoption -- let us review some of the key trends and predictions in this space.
Adapt Now or Pay Big for Downtime
The numbers alone demonstrate why adoption of DevOps has become mainstream. As more business is generated by technology, companies that are not changing business models in the face of disruption see immediate financial implications.
As reported by IDC, the average infrastructure failure at a Fortune 1,000 company costs $100,000 per hour. For a critical application fail it can cost upwards of a million dollars an hour – and the average cost of downtime per year ranges between $500 million and $2.5 billion per year.
On the other hand, high performers who employ DevOps principles saw 200 times the number of deployments than average or low performers, according to Puppet’s 2016 State of DevOps report . They also saw 24 times faster mean time to recovery, 300 percent lower change failure rate, 2,555 times faster lead times to delivery changes.
Perhaps the biggest differentiator from the IDC report on organizations who failed to change, was that the cost of downtime was significantly lower for high performers, averaging $37,500 per incident . When deployment fails in a DevOps environment, the smaller batch sizes are easier to fix and roll back, enabling faster response times.
The consequences of downtime and disruption have grabbed the attention of the C-Suite. IDC reported that by the end of 2017, two thirds of CEOs at Global 2,000 companies will have digital transformation at the center of their corporate strategy. This will drive more adoption of DevOps to enable applications and technology solutions that power our business to be delivered to end customers faster and with better quality. This mandates that silos are broken down, and the delivery pipeline be streamlined - end to end - and be treated for what it is: a key differentiation for businesses today.
At the enterprise, DevOps adoption tends to have specific needs, such as legacy systems, compliance and regulations, hybrid architectures and growing security risks. Software development and delivery must also be supported at a massive scale, so the businesses tend to desire technology solutions that have been proven and can address the needs of the enterprise.
ARA Gets the Most Bang for the Buck
Application Release Automation (ARA) solutions continue to grow, and prove that they offer incredible value because they accelerate the release and improve the cycle times of application delivery.
Gartner expects 50 percent of Global 2,000 enterprises to adopt application release automation by the end of 2020. In 2016, they estimated that 10 percent of enterprises had adopted ARA. Essentially, the expectation is that 40 percent of the Global 2,000 will be adopting ARA over the next two and a half to three years.
Microservices and Containers Rise in 2017
In 2017, the adoption of microservices has grown significantly. Microservices are suited for DevOps because of their isolated nature, well defined interfaces, and ability to improve cycle times without the risk of breaking the broader system. Containers are a compatible deployment vehicle for microservices because they're designed to run one process with minimal deployment and run time overhead.
The container market is also a dynamic and quickly changing. Docker has become the standard for developing containers, but in terms of the run time, the clustering and the orchestration environments, there's a lot of flux in the market. The ideal solution would be future-proof in that it will allow users to easily switch from one platform to another - as the needs of the business change. If the enterprise is leveraging microservices and containers, it is increasingly important to incorporate solutions that support containers and microservices as a first-class citizen along with supporting the traditional application models as well.
Always Shift Left
DevOps tools and solutions continue to accelerate the notion of shifting left. Automated testing and performance monitoring is done earlier in the lifecycle to identify performance degradations early on, service virtualization allows the enterprise to emulate services for improved application testing, and security verification is now a part of the delivery pipeline. DevOps orchestration tools automate the entire pipeline from end-to-end enabling comprehensive traceability. This is an advantage for compliance, as it can be innately built in through end-to-end automation.
As enterprises leverage automation, demand is increasing for integrations with best of breed tools. We can expect to see more strategic alliances between the tools and solution providers because the enterprise increasingly seeks out of the box enterprise grade integrations.
Assess Your Pipeline and Release Health
As tools are tied together and the pipeline is orchestrated, the massive amount of data being captured is actionable. The use of predictive analytics will pick up, and we will see the application of machine learning to get insights into the health of your delivery pipeline. Predictive analytics can reveal the risk of a pipeline failing and can be used to determine the health score of the release.
Culture is the Clear Differentiator
While we need to pay close attention to these trends in tools and technology in order to deliver faster high-quality software, culture and process remain staples of the DevOps journey.
Culture is the top prerequisite to successful DevOps. An organization can have all of the automation tools, but without establishing the organizational requirements and addressing the cultural issues, the outcome is not as impactful. Culture is the key to breaking down the organizational silos, and fostering collaboration across all the groups that need to work together to deliver a business outcome.
ITIL will Adapt for DevOps
ITIL will need to adapt for this new accelerated and automated world of DevOps. Automation can optimize for speed and enable pre-approved changes upon passing an automated test. As a result, CAB will move to more of an auditing function. There will be more discovery of Continuous Integration (CI), and as the software delivery pipeline is creating CI artifacts, those will be automatically populated into the Configuration Management Data Base (CMDB).
DevOps at this point is no longer a fringe movement. It's become an existential necessity in many cases. Application release automation has been identified as the highest value technology for DevOps adoption. Containers and microservices are gaining momentum and are suited for the DevOps approach – but they need to coexist with the traditional architectures. The shift left is growing as it identifies quality, performance and security issues sooner.
Culture is the key differentiator for DevOps. If processes are not optimized for speed, the automation will not be successful. When culture and processes are in alignment, the automation tools will be more effective in helping organizations adapt to disruptive industries.
As part of the upcoming Jenkins World, held August 31 in San Francisco , I would be participating in a video panel discussion, hosted by DevOps.com, with Cloudbees' CEO Sacha Labourey, Rob Stroud from Forrester, and others. I will be discussing some of these trends and other DevOps predictions to pay attention to, as well as wrapping up the show.
If you're attending Jenkins World SF, be sure to meet us at the CloudBees booth, and stay tuned for the video of the recorded panel after the show!