Ever since it first emerged in the late 2000s, a DevOps approach has helped countless organizations become more secure and agile. It ushered in a raft of changes, most notably promoting greater collaboration between separate and often fragmented teams working across the likes of security, infrastructure, development, quality assurance and support. Today, we’re on the cusp of further transformation, this time concerning the way IT organizations handle modern workloads from the edge to the cloud.
Considering the benefits DevOps provided, it’s important to note that beyond helping individual teams gain a better understanding of the needs of other departments and, in turn, develop better and more secure products, it also demonstrated the need to test assumptions and think carefully about how to manage applications in modern architecture, and how to best leverage engineering processes for maximum benefit.
This process worked, but only so long as applications weren’t required to meet ‘real-time’ demands. As more and more organizations now lean toward an ‘always-on’ business model to provide anywhere anytime access – likely as a result of consumer and business expectations amid the pandemic – many have also shifted to the cloud to gain even greater agility.
Despite the rise of DevOps, there remains a divide between development and operations teams today. And alongside this, the accelerated pace of the relatively recent ‘move to cloud’ has also accelerated the need for this challenge to be addressed. Given their ‘always on’ nature, cloud workloads are constantly being deployed, run, developed, and updated, requiring an integrated approach to managing collaboration between application engineering and operations.
In this context of the cloud, we should not treat engineering and operations as separate activities. In fact, engineering-led operations is the first big step towards moving any organization to a point where operations don’t exist independently.
The vision behind engineering-led operations is simple. Wouldn’t it be incredibly useful to automate mundane tasks, allowing engineers to automatically monitor and address challenges? Or perhaps a code that can fix problems itself with the correct actions as soon as an issue is detected, or even before it becomes a problem.
Premkumar Balasubramanian is Senior Vice President and CTO of Digital Solutions at Hitachi Vantara.
Embracing something new
A widespread shift towards engineering-led operations would bring greater agility to any cloud environment. Today, most organizations struggle with engineering a successful cloud setup. As many continue to grapple with the complexity and challenges that arise with cloud management, engineering-led operations can step up to the plate.
Let’s explore how an engineering-led approach might look in practice.
Today, development teams often have a features backlog, which everything goes into. All feature requests are competing, regardless of their impact on an organization's bottom line. The issue this presents is the possibility of a feature critical to a workload’s stability getting lost in a significant backlog. While this stability feature indirectly impacts revenue, there’s a possibility it would be deprioritized in this circumstance to favor a feature identified as revenue generating.
If organizations adopt engineering-led operations, appropriate teams first agree on service level objectives (SLO) and error budgets. If these variables were to then exceed allowable norms, both teams could seamlessly deprioritize new feature requests and instead shift focus on reliability-related features until production is stable.
Engineering-led operations is all about bringing infrastructure, applications, and the data closer together. Unlike DevOps, the approach is not just concerned with agility to operations. It also speaks to the integration of these two functions with shared goals and ways of working.
In order to build modern businesses, it’s imperative that this trend toward function integration has the investment to grow further, allowing modern software engineering to become less about simply building products and services and more about championing productive and reliable business.
Boosting reliability in cloud
For ‘always on’ businesses, the challenge of running and sustaining workloads efficiently has become the true test for success or failure. In the modern age, when glitches arise, outages have the power to threaten both credibility as well as revenue.
No modern business wants to trend on social media for a cloud-related outage, and every highly reported event should remind us that the cost of being unreliable is steep. When Google Cloud and Oracle servers suffered cooling-related outages during the UK’s record 2022 heatwave, it caused issues for individuals and organisations using services to host their websites. By the time the issue was resolved at 10am the next day, thousands of comments relating to the issue had been posted across various social media platforms.
An integrated, engineering-led approach helps prepare and even avoid situations like this. It creates the conditions that allow teams to stay informed, poised to act, and assured solutions align with a set of shared goals and practices.
In the fast-paced world of “everything-as-code”, businesses urgently need to keep up. Failure to do so could mean the difference between market leadership and falling out of the competition.
