DevOps in 2019: Gene Kim's Trends and Predictions for Next Year
It's hard to believe another year has come and gone, but here we are, on the doorstep of 2019.
The Accelerate State of DevOps Report for 2018, DORA's annual assessment of the software delivery landscape, noted that the number of DevOps teams qualifying as high-performing grew to the point where it justified adding an additional tier categorized as elite. So, does that mean we've hit the peak for DevOps heading into 2019? Or are we really just getting started?
Enterprise software delivery isn't a "what have you done for me lately" game; it's all about what's next. To kickstart 2019, we asked DevOps thought leader Gene Kim and XebiaLabs' VP of Product Development, Andreas Prins, to break out the tarot cards and give us a reading on where they see the industry heading in the New Year.
You can watch the full webinar here. Or check out Gene and Andreas' six crucial DevOps focal points for 2019 below.
1. The performance rate of DevOps teams will continue to improve in 2019
The Accelerate State of DevOps Report categorized 48% of their respondents as high performers. This high rate suggests that the overall industry is improving its software development and delivery practices. It also indicates that a high rate of DevOps performance is more easily attainable than it has been in the past and is no longer something reserved for an exclusive set of teams.
2. Cloud and microservices will increase the number of Continuous Integration pipelines and accelerate CI cycles
In 2019, as teams continue to find ways to accelerate delivery, we'll see the number of CI pipelines go through the roof. On top of that, Cloud and microservices will be a big driver in increasing the frequency of CI cycles because the unit of deployment will be smaller, and the underlying platform will change often. The ability to integrate CI pipelines into the larger DevOps pipeline will be essential for preventing bottlenecks as the number of CI pipelines increase and their cycles accelerate.
3. Tech leaders will look outside of the traditional technology value stream
The DevOps Enterprise Summit, the annual industry event hosted by Gene's company IT Revolution, gives him a unique look at the leadership challenges facing today's digital executives. The biggest roadblocks these execs are encountering are now outside of the traditional IT scope, in areas like project management, information security and compliance, and business leadership. To help bridge this gap, the DOES event puts tech execs on stage with their business counterparts. Understanding the needs of the business, regardless of where they lie, and demonstrating the value software delivery brings to these needs, will be the next frontier of technology leadership.
4. DevOps as Code will be key for accelerating the creation of value
DevOps as Code allows teams to define everything in their DevOps pipeline as code--deployment packages, infrastructure, environments, release templates, dashboards, and more--and maintain these definitions in source control management. In 2019, DevOps as Code will become a much bigger driver for accelerating value, with benefits such as simplified cloud migration, easier on-boarding of projects, pipelines, and teams, and standardized security and compliance built into a fully auditable pipeline. GUI's won't go away. On the contrary. Having both DevOps as code for technical team members along with a GUI for less technical ones will be necessary for organizations to extend DevOps across the entire organization.
5. Next-gen infrastructure will increasingly rely on functional programming
Functional programming will increasingly become part of next generation operations. Functional programming languages are easier to use and less error prone than imperative languages because they’re declarative, immutable, and recursive. They’re also aspect oriented, allowing for changes to be made across modules and components without having to touch each one. The techniques of functional programming are already affecting how IT infrastructure and operations are done. For example, immutable infrastructure is at the core of Docker, and sidecars in Kubernetes are a form of aspect-oriented programming. Functional programming enables infrastructure and operations teams to build platforms for developers so that developers don’t have to do it themselves, which is important for freeing them to focus on creating features.
6. Smart analytics, machine learning and AI will be increasingly embedded in the DevOps process
As organizations continue to struggle with processing huge amounts of DevOps data, we'll see a trend from simply doing "point views" of data through individual tools, to using smart analytics, machine learning, and AI to harvest data and learn from it. Smart analytics--essentially reports and dashboards that allow people to make decisions about releases--will continue to be important. Companies will also start embedding machine learning in the DevOps process, which will allow for automated predictions about whether a release will make it to production. AI will enable teams to go a step deeper, for example, automatically predicting the most high-risk tasks so that teams can make point decisions about releases as those releases move towards production.
A Resolution You Can Keep in 2019
If 2019 is the year you're vowing to release better software faster, XebiaLabs is here to help. DevOps at enterprise scale requires a lot of tools and includes a ton of moving parts. To succeed, enterprises need to ensure the right things are happening in the right order, with the right set of controls and processes in place. The XebiaLabs DevOps Platform manages and orchestrates all of the tools and processes in the overall release pipeline, from planning to production. With XebiaLabs, you can automate and synchronize the many interdependent tools and steps involved in deploying applications to various environments. Try it today for free!
More from XebiaLabs
- 2019 DevOps Predictions: Gene Kim Places His Bets for Next Year
- DevOps Hits Adolescence – what’s next?
- The Reality of Software Releases
- Going Code-native with DevOps as Code