You are here

Why continuous assessment is key to scaling enterprise DevOps

public://pictures/John-Jeremiah-Technology-Evangeslist-HP.png
John Jeremiah, Evangelist, GitLab

DevOps isn't new, isn't fancy, and certainly isn't a silver bullet. DevOps is over-hyped, over-marketed, and oversold as the latest and greatest way to "fix" IT.

It's also the key to your future as an IT leader and to your organization's future survival in the market. If you can't figure out how to streamline and accelerate innovation through a relentless process of continuous improvement, your company risks joining a long list of former industry leaders, from Blockbuster to Borders to RadioShack, that were unable to react quickly to market disruptions.

[ Get Report: The Journey to Enterprise‐Scale DevOps: Becoming DevOps Determined ]

A solid foundation

DevOps evolved from a rich history of process improvement. You can trace the roots of DevOps to Deming's process improvement efforts and his teachings on statistical process control and the concept of using measurement and metrics to drive improvement. When I look at the foundation of DevOps, I see the principles of Colonel John Boyd's OODA loop, which form the foundation of modern military tactics and focus on rapid decisions based on observing what's happening in the real world. There's also a clear linkage between DevOps and the theory of constraints from Goldratt, as well as the principles of lean manufacturing from Taiichi Ohno's Toyota Production System. DevOps is incredibly powerful and transformative because of its deep roots and rich legacy.

Feedback and measurement is a common thread that weaves through all previous business improvement techniques. Data-driven decisions and focused improvement lead to increased quality and efficiency, and the use of feedback to accelerate delivery makes DevOps a disruptive IT concept in much the same way that lean manufacturing disrupted the automobile industry. Popular blogs often describe DevOps as dependent on culture, automation, measurement, and sharing (CAMS), the four ubiquitous themes in the DevOps movement. In the enterprise, however, measurement takes on an important role in helping to enable and support the ongoing change and innovation DevOps fosters.

In a recent keynote at Forrester's "Forum for Technology leaders", analyst Michael Facemire described the importance of having a feedback-driven lifecycle. Modern application development must be based on building a minimally viable project and then seeking customer feedback to validate utility and the application's value. By using feedback to shape the future direction of the application, you let the customer tell you what they want and don't want.

In the context of DevOps, however, Facemire doesn't go far enough. Done right, DevOps is a feedback development lifecycle where assessment and feedback are the foundation and are embedded in every step. It forms a series of feedback loops that provide immediate insight and responses to each preceding step in the process and deliver feedback from customers into the application's business value. This feedback process, which must be automatic and continuous, is called continuous assessment. It's the processes and framework that enable the business and its developers to measure and track defects, build issues, performance issues, security issues, business value, and the use of key functions. Continuous assessment drives tactical decisions about quality and rework, as well as the strategic investments into future product releases and updates.

[ Partner resource: Take Security Journey's first two white belt modules for free. ]

Scaling through continuous assessment

In the enterprise, continuous assessment is a cornerstone of scalable enterprise DevOps. It empowers development and operations teams to accelerate delivery and iterate rapidly based upon accurate and actionable feedback about their products.

Here are three recommendations you can use to drive home the value of continuous assessment in your organization:

  1. Incorporate monitoring tools and processes as early as possible in the development of your application. Monitoring is about more than just production performance: You must be aware of the health of the application at all stages of development. By monitoring earlier in the process you will gain valuable insight into the application before it goes to production, making it easier to correlate and compare how the application behaved in test environments and in production.
  2. Measure to determine which features in your application are most valuable and which are at least valuable. Many features and functions are never used. They're expensive to maintain and make innovation more difficult, effectively slowing down your team.
  3. Figure out how to measure and track the business value that the application enables and delivers. This information is crucial when business and IT leaders prioritize future investments and make the inevitable trade-offs that arise. Rapid innovation in this part of your business could be a competitive advantage that disrupts your competition.

Everything is an experiment of sorts. Assessing the outcome and impact is a critical discipline that IT must embrace in its journey to implement DevOps. The process must be automatic, continuous, and comprehensive. Without continuous assessment in your DevOps toolkit, you run the risk of flying blind and missing huge opportunities that are easily within your grasp.

[ Webinar: Paving the Last Mile to Production: Putting the "O" in DevOps ]