Adaptavist's DevOps focus for 2024

For us at Adaptavist, 2023 was a year where generative AI hit the big time, and we started to assess how to make the best use of a ground-breaking new technique. Early aspirations suggested that by now, we would have improved developer productivity by having all coding written by a Chatbot, but in 2024, we'll see a more nuanced approach. Whilst the panacea of AI writing complete and correct code every time is a way off, integration of chatbots into IDEs is already beginning, allowing developers to exploit machine-learned hints to write and improve code. Expect to see the rise of generative AI in a DevOps context, with teams exploring the use of AI and ML in many areas.

Testing and QA

Automation has already helped speed up testing, making it more repeatable and scalable. And well-integrated AI tools take that automation one step further. Writing unit tests is a sometimes repetitive but vitally important task that can be overlooked. This makes it an ideal application for generative AI to write cogent and relevant testing code that improves test coverage. This will allow feedback to be more embedded in the development lifecycle as these tools become available and more widespread in IDEs. By analysing test data and identifying patterns, AI tools can create new, more effective test cases that are targeted towards areas where issues are more likely to arise.

Monitoring and observability

While it's fairly straightforward to see if a service is 'up' or 'down', contemporary tooling efforts are using machine learning techniques to detect anomalies in performance data, which could be indicative of a potential problem. AI offers a data-driven solution rather than teams relying on experience and their own intuition. It analyses past patterns, looking out for trends that might otherwise go unnoticed and making intelligent predictions about upcoming demand so that you can avoid over or under-resourcing.

With AI, you'll also get real-time insights that traditional monitoring methods would otherwise miss; streamlined incident management, with incidents prioritised based on their severity; and can save time and resources identifying root causes. We'll see organisations using AI and ML to analyse log data alongside other metrics to tell you where problems lie.

The rise of cloud computing has continued in 2023, but with the growth of this multi-billion dollar industry came some increased concerns over costs. Cloud computing is a huge operating expense for most businesses, but the variable spend model means it's tough to avoid cloud waste and make significant savings. For example, the founders of Basecamp made a highly publicised move away from cloud computing and back into the data centre. While we don't expect this cloud repatriation trend to grow hugely, it demonstrates increased rigour around costs, a concern for anyone using cloud computing at any scale.

Tooling again will evolve in this segment. Older cost management tools rely heavily on tagging to allocate cloud costs, but the adoption of tagging has proved difficult in many businesses, especially when trying to allocate the costs of shared resources. Tooling to analyse these costs are evolving – and we'll see more of this as businesses reconcile a finance approach of 'spend as little as possible' with understanding the return on investment of cloud costs in more depth. The recent emergence of FinOps as a defined practice increases our shared understanding of the nuance in this area.

When used alongside best practices and a FinOps culture, platforms like CloudZero give everyone the information they need to manage cloud costs proactively. Effective cloud spend, supported by cross-functional decision-making, can drive revenue, enable you to increase product release velocity, gain a competitive edge, and ensure everyone knows why investment is happening (and why it's not).

Platform engineering plays a pivotal role in the realm of DevOps, offering a foundational infrastructure that accelerates software development and deployment and solves the problem of cooperation between software developers and operators.

It's a cliche to describe freshly titled platform engineers as the people who were named site reliability engineers last year and DevOps engineers the year before that. But much as the SRE job role conveys added meaning, rather than a simple rehash of 'DevOps engineers', there is an evolution in progress as platform engineering becomes a practice in its own right.

In the contemporary DevOps landscape, platform engineering teams are increasingly providing ‘plumbing’ as a service, delivering the fundamental components and infrastructure that enable seamless development and deployment to happen. This approach allows development teams to focus on their core responsibilities without being burdened by the intricacies of infrastructure management.

In 2024, we'll see more thoughtful definitions of what platform engineering means to teams practising DevOps as larger organisations build out distinct capabilities and service needs that can't be accommodated by off-the-shelf cloud products alone. Expect to see more focus on product-centric developer self-service. A key point here will be how internal teams collect requirements and feedback from internal customers, as platform teams focus on adding tactical value rather than replicating services that could be bought in from elsewhere.

Expect to see organisational design changes, with more stream-aligned teams as promoted by Team Topologies emerging. Engineering the plumbing to make those teams effective will increase independence and efficiency within these specialisations, promoting a more streamlined and collaborative environment.

The emergence of standards within platform engineering has become a critical aspect of this discipline. According to Gartner®' by 2026, 80 percent of large software engineering organisations will establish platform engineering teams as internal providers of reusable services, components and tools for application delivery. Platform engineering will ultimately solve the central problem of cooperation between software developers and operators.’1

High-performing organisations understand that focusing on developer experience – providing the right tools and processes for developers to thrive – is key to delivering quality software reliably and quickly. Almost three-quarters of businesses improve their productivity by enhancing DevEx, with other effects including revenue growth and improved customer satisfaction.

Ecosystems providing internal developer platforms will continue to evolve into products such as venue.sh, as DevOps teams take the learnings from frameworks like Backstage to make developers' lives easier. Teams who want to innovate will start exploiting new techniques. A great example of this is using the IDP to quickly start new coding initiatives with the help of agreed-thought-out skeletons for cloud infrastructure. Another is using a service catalogue to provide ready access to available microservices. These teams understand that providing a frictionless route to writing and deploying code is a key requirement.

Cloud development environments offer scalability and collaborative capabilities, but developers often crave the immediacy and feedback loop provided by local environments. So, how do we bridge this gap?

Expect to see some further acceleration of development environments moving towards the cloud. Developers are used to a local development environment that is responsive and customised for them, and innovation in providing web-based IDEs will begin to deliver this in the cloud. Deeper integration with ancillary services used during builds, such as security scanning and generative AI integrations, are made easier by running in the cloud. The ability of centralised teams to leverage tighter control of these environments for security and audit purposes will lead to more adoption in this area.

Furthermore, the gap between build environments and deployment environments will narrow, with services such as Gitpod and GitLab offering full cloud development environments directly linking to Kubernetes clusters for easy and standardised deployments.

Despite the push towards the cloud for many development tasks, local development tools continue to improve and gain traction. Tools like Telepresence – which allows developers to test in a production-like development environment by connecting the copy of their service locally to their remote dependencies – will be key here. They can help with the transition to cloud development by reducing friction and maintaining the crucial rapid feedback loop that local development offers.

There's plenty more where that came from. Do you want to pick our brains about what's to come in DevOps and other software development practices? Need support to get set up with new tools or techniques? Are you ready for a DevOps boost in 2024? We're here to help.