New year, new me, new DevOps? Not quite, though the new year is bound to come with a few changes. If the 2010's were the decade software development really took over the world, the 2020's will be the decade in which automation and DevOps practices shine. But with everything going on, it's sometimes difficult to keep an eye on the bigger picture. This is why we at Solidify every year try to look forward and predict how DevOps will transform and what upcoming key topics might be.
In 2018, we talked about DevOps metrics and the value of leadership in DevOps. Last year, we mentioned Ops for Devs, DevOps at scale and how important compliance would become. In this blog post, we dive into some of what we believe to be key topics for 2020.
Site Reliability Engineering, or SRE for short, is a relatively new concept within DevOps. Just as with DevOps itself, many people and organizations have different interpretations and definitions of the concept. However, many consider Google's Ben Traynor to be the founding father of SRE, and he describes it this way:
It’s what happens when you ask a software engineer to design an operations function. So SRE is fundamentally doing work that has historically been done by an operations team, but using engineers with software expertise, and banking on the fact that these engineers are inherently both predisposed to, and have the ability to, substitute automation for human labor.
We strongly agree with Ben in that operations can, and should, be handled similar to traditional software development. Utilizing automation, tearing down silos, collective and shared responsibility - all of these are familiar to those working with DevOps, and it's the same when we talk about SRE.
In short, we see SRE as a great complement to DevOps. Expect to see and hear a lot more about SRE this year.
2. GitHub everywhere
Open source software has grown intensively the last couple of years. Many organization and companies put their source code out in the open and invites people to contribute and experiment on existing code. This is true in the DevOps world too - container orchestration platform Kubernetes is open source, and open source is also the basis for the container platform Docker. There are many more examples, but the point is that open source is here to stay.
We believe that humans want to collaborate. People want to work together, and GitHub is currently the best option to do so in regard to software development. With a bigger code base and number of open repositories than ever, GitHub is the largest open source platform in history. Furthermore, we see an increasing number of companies centering their development around a GitHub powered core and developer communities.
In addition to open sourcing, so called inner-sourcing is also growing in popularity. This type of development, effectively involving a model of open source but kept within company borders, enables companies to live up to values such as transparency and collaboration while still maintaining control.
And if you're still not convinced and want to keep things on premises? Well, GitHub Enterprise offers just that, with all benefits of github.com.
3. DevOps for data science
Data science is perhaps one of the hottest topics in the IT world right now. As the world and us in it create more and more data, it's become apparent that we need to think hard about what to do with that data. Everything from machine learning to artificial intelligence and simple data evaluations are influenced by data science. It touches almost every industry and is surely a part of the future of IT.
So, it follows that data science will need to become efficient. Just as with regular software development, data science can draw use of DevOps principles and values. For example, a data scientist working on a machine learning model for making business crucial predictions would most unquestionably benefit from an efficient CI pipeline. Or perhaps automating tests to make sure that the model itself isn't deviating in undesirable ways.
The recurring need for compliance and security likely won't go away. DevSecOps incorporates traditional DevOps principles and methods with the aspect of security. It includes everything from ensuring compliance to relevant regulations and laws to data integrity and encryption.
As the IT landscape keeps shifting beneath our feet, it becomes more and more important to make sure we stand on solid ground. While the popular belief is that DevOps increases risk, the fact is that the opposite is true. Through integrated testing in both build and release pipelines, security and compliance requirements can be satisfied from day one and forward. Concepts such as shift left, security as code and cloud transformations all contribute to classic DevSecOps advantages, such as early error detection, greater speed and agility, and shorter time to recovery.