Today’s customers expect a constant stream of ‘new’ applications and digital service experiences, and so companies everywhere are transforming into software businesses to meet those demands. This trend has an even greater importance now, as the world battles with the current crisis, with digital services the primary link many of us have with the outside world – both professionally and as consumers. Innovating at the rapid speed needed to meet these demands requires agility within IT operations, and almost all organisations are using the cloud in some capacity to achieve this. However, the cloud’s dynamic nature has also led to a surge in complexity, with 76 per cent of IT professionals stating this as the biggest barrier to productivity among operations teams.
IT teams spend enormous amounts of their time piecing together metrics and alerts to ‘keep the lights on’, as they struggle to manually capture everything happening in their IT environment using traditional performance management approaches. This massively eats into the time IT teams could be spending developing and delivering new, value-added services to the end-user. It’s also especially challenging in the current situation as the majority of business and IT teams work remotely and are stretched more than ever. However, an automated approach to IT operations, known as NoOps, offers an attractive alternative for IT leaders, enabling them to run IT operations autonomously, so remote IT teams can develop and deploy new functions and services much faster and with far less friction.
The road to NoOps
NoOps is the concept of an IT environment whereby the use of automation and AI-assistance radically reduces operations staff. While this is especially attractive in the current crisis, as it reduces the need for employees to be physically present, NoOps is also crucial to improving IT operations in the longer term. Essentially, it’s driven by looking at what could potentially go wrong, taking steps to proactively prevent that through automation. However, this can only be achieved if businesses have a solid CI/CD toolchain in place with AIOps fully integrated into their ecosystem. With this approach, AI is used to analyse and triage monitoring data at a higher volume and faster speed than could ever be achieved manually. This uncovers precise answers and detailed performance insights, in real-time, creating a stream of software intelligence that makes sense of the endless alerts. This can then be used to trigger the automated responses that are at the heart of NoOps.
However, baking in automation and self-healing to create a continuous delivery process has some – including Mike Gualtieri, the Forrester analyst who coined the term NoOps – speculating if this will spell the end for DevOps. Their argument is that NoOps eliminates the need for developers to collaborate with operations, which may lead to a decline in ideas that drive innovation and help to maintain seamless user experiences. Unsurprisingly, this theory can lead to resistance from DevOps teams to embrace the NoOps approach, as they worry they will become redundant.
Evolution, not revolution
Contrary to what its critics profess, NoOps is far from the ‘end’ for DevOps and more akin to its next natural evolution. With DevOps, operations teams apply development practices such as version control, scripting and automation to address potential performance issues. With NoOps, it’s like the inverse, as developers begin to think like operations teams. The result is operations teams can work in tandem with developers towards the common goal of driving innovation for the business and its end-users. DevOps teams will therefore no longer find themselves working at half power towards the goal of creating new services, as the other half of their team is occupied with ‘keeping the lights on’.
For organisations that achieve NoOps, it will be the biggest transformation of software delivery processes since the emergence of DevOps. We’ll begin to see DevOps evolving to align closer to the needs of the modern business, as organisations embark on the journey to autonomous cloud operations. This is far more suited to the current climate, where stretched IT teams must achieve a faster pace of innovation, as problems are fixed automatically in the development phase, speeding up the delivery of new software experiences to the business and its customers.
Building the case for NoOps
While organisations may recognise the benefits NoOps can bring, it can still be difficult for them to get DevOps teams on board. However, having the full support and commitment from those who will be involved in the shift to NoOps is crucial to success as it requires a fundamental transformation in how teams think and operate. Giving DevOps teams a more concrete idea of how NoOps will benefit them can dispel concerns that there will be no more need for collaboration between developers and operations.
Businesses should therefore embark on an education process to ensure teams are clear about how their roles will evolve, they’re comfortable with the tools that will be available to them and are happy about working with a shared goal in mind. For developers, this involves informing them of how NoOps can remove bottlenecks, as they won’t have to spend time in a cycle of debugging to figure out where things went wrong with their code. For operations teams, IT leaders should highlight how NoOps can help them elevate their role within the organisation and take a more active role in driving innovation. Operations teams will no longer have to spend time on tasks that simply ‘keep the lights on’, instead they will be focussing on value-added activities – such as continuous deployment and innovation.
Reaping the NoOps reward
As the scale and complexity of the cloud continues to grow, and organisations are also now busy working through the unprecedented situation the current crisis has created, businesses need to find a way to help their DevOps teams sooner rather than later to make sense of their IT environment and NoOps is the key to this. Those who implement it successfully will be able to supercharge innovation and deliver new, high quality services to end-users much faster than ever before.
Michael Allen, VP & CTO EMEA, Dynatrace