DevOps “isn’t just for startups anymore, as more and more enterprise companies have realized they require agility to stay current and modern,” said Ben Thrift, director of engineering with Dell Boomi. “New projects are being built with DevOps in mind, and tooling and automation are making it easier to retrofit legacy projects with DevOps principles.”
Furthermore, DevOps is about more than speed. “It’s about using speed to make the applications more intelligent and deliver optimal functionality that generates improved business outcomes,” Carr pointed out. “DevOps is supposed to help agile teams ‘fail fast,’ thereby enabling them to perform rapid or even simultaneous A/B testing, and so forth. Implicit in this is the use of far more data and analytics.”
RAPID DEPLOYMENTS
Other initiatives moving organizations toward greater agility include deployments of microservices and containers, which enable applications, services, code, and even data to be moved or reused in the most expedient technology environments, such as cloud platforms. However, there is also still much work to be done in these areas. Most enterprises have made some transition to DevOps, containerization, and microservices, but long-established corporations are struggling to turn around their processes and methods, said Venzl. “New projects tend to adopt agile practices faster, benefiting from the clean-slate state over already rolled out projects running in production.”
This all means that “it is a brave new world that is unfamiliar to many, so there are more questions than answers today,” said Brey. “While progress is definitely being made, especially in markets where the level of competition is high, the access to agile techniques is ready, and there are lower barriers to change.”
Julie Furt, senior vice president, worldwide consulting and training services at MarkLogic, sees “pockets” of progress with these next-generation technologies, but often it’s simply lip service. “Some organizations pretend to be agile, while really just putting a wrapper on their traditional approach—or worse, have interpreted ‘agile’ to mean that you can figure out what you’re building at the end, not establishing clear goals and then iterating toward them.”
Todd Moore, VP of open technology and developer advocacy at IBM, has seen such a radical transition at his own organization. “Front-ending hundreds of gigabytes of data that lives both in the cloud and enterprise systems with APIs just goes with the territory these days,” he said. “Microservices can automatically render that data into content and interactive services and this is exactly the case for one of IBM’s most important and visible websites, IBM Developer.”
There is also a risk seen with adoption of these new technologies and practices—an assumption that they alone will set organizations on the right track. “A common misconception I have seen with people who are new to Kubernetes, is that the Kubernetes orchestration layer and its self-healing properties will solve all of your problems,” Frank Reno, senior technical product manager at Sumo Logic, said. “While there are many problems and headaches it does solve, you cannot abandon your core monitoring principles. It is still critical to monitor the orchestration layer, in addition to all your applications, to ensure the orchestrator is orchestrating. You need a robust, scalable monitoring solution that can scale with the volume of data that comes from running applications inside of Kubernetes.”
IMPACT ON DATA ENVIRONMENTS
The database space in particular has been a challenge in moving to more agile approaches, due to greater centralization of “both system and application DBAs,” said Jones. “Any type of request to the DBAs could take days, weeks, or even months. In order to improve agility, the developers need to be in control of their databases."
To fully embrace an approach such as DevOps, Jones continued, “talented DBA teams need to become passionate about helping developers take control of what the DBAs used to do. The system DBAs are now helping to provide automation to the development community. They do it in a way that allows developers to securely deploy databases and connect their applications to them in less than an hour. This is something that in large organizations used to take weeks or months. The application DBAs are now either on development teams or working with them to modernize their applications. Part of this modernization is allowing teams to use schema management libraries in their code to deploy schema changes. They are also teaching teams how to effectively and efficiently monitor and maintain their databases.”