What will be the key DevOps trends for the year ahead?
As boardrooms wake up to the race for digital transformation and their consumers demand a new digital journey, software development teams need to step up to the challenge.
The last 12 months have seen further innovation in DevOps, along with the acceleration of cloud adoption and advances in technologies, such as containerisation and microservices.
Automation and AI
No prizes for my first prediction. Automation technologies will continue to extend further into the DevOps pipeline to deliver a quicker time-to-market with reduced errors, freeing up staff to focus on innovation as well as cutting costs. Automation will also become more prevalent in areas such as predictive software development.
Artificial intelligence is already starting to facilitate predictive analytics and coding to replace manually intensive tasks with intelligent insights, recommendations and automation. And as more testing is required to identify and rectify performance and security issues in production, the use of test automation and test-driven development will also increase.
Continuous monitoring of application use and performance will power feedback loops to highlight and address problems instantly.
Competition drives consolidation
2018 has seen industry consolidation with the acquisition of GitHub and Red Hat, as the big vendors bolster their DevOps propositions and aim to become the go-to ‘one-stop-shop’. No doubt we will see more of this strategy in 2019 as companies strengthen their position with customers and reduce integration headaches between tools.
But as the software development and DevOps communities are well known for getting behind popular tools, new leaders have emerged. It is likely we will see more from the large cloud providers as they entice developers to their platforms.
Containerisation and microservices
The growth in microservices and distributed systems has driven a massive increase in container workloads. Running containers in production is becoming standard practice and Kubernetes has been widely adopted as the container orchestration tool of choice, evidenced by the launch of Azure AKS and AWS EKS.
We will see legacy platforms start to become containerised or replaced as the benefits of this technology are now accepted. As organisations move from monolith to microservices, serverless computing – or Functions-as-a-Service – offered by leading cloud providers, will be used increasingly to focus on business logic rather than infrastructure.
Adoption will accelerate as organisations realise they can deliver applications at record speed and lower cost and this trend will undoubtedly generate increased workloads on cloud platforms. However, there is a risk of locking in applications to platform APIs and other services, which will be a concern for some.
The increasing use of microservices and distributed containerised systems bring more pressure on the network, which must be as agile and automation-ready as the applications running on it, rather than create a bottleneck.
One of the technologies for 2019 that promises to deliver high-performance, failsafe, dynamic and flexible networks but without the cost of legacy MPLS, is SD WAN. We will also see a growth in multi-hybrid cloud adoption as it no longer becomes simply a choice of cloud provider, but a means to optimise the cost, performance, architecture and location.
DevOps at scale
As business executives focus on speed and competitiveness, successful DevOps initiatives will drive more focus and investment. We will see DevOps scale to more teams and projects, with additional governance as businesses shift to enterprise tooling. Open Source has been very popular and an effective way to prove the capability of tools to automate the DevOps pipeline, but enterprises now need the assurance of stability, security, scalability and compatible integrations as this becomes a critical function.
Security and DevSecOps
There are over 100 billion lines of new code being produced every year – and each one can introduce a new vulnerability. Standard security checks are not enough anymore, especially with increasing audit and compliance requirements.
Security is increasingly shifting towards developers, to ensure applications are secure by design and vulnerabilities introduced by the increasing use of open source components are captured and addressed. If the right tools are not in place to identify vulnerable software, attackers will find and exploit these weaknesses first.
Marlene Spensley, application optimisation practice lead, Nuvias