An important prerequisite to improving something is to start measuring it. The same applies to the DevOps maturity of your teams. It is essential to identify the areas of improvement, understand progress, and track the effectiveness of any process or change rolled out as part of DevOps practices in your organization. But what exactly needs to be measured, you may ask?
Accelerate 2022 State of DevOps Report (Ref:https://services.google.com/fh/files/ misc/2022_state_of_devops_report.pdf) mentions two clusters of analysis, one targeted toward software delivery performance , and the other toward operational performance. These reports are based on inputs from more than 33k respondents from different organizations. Figure 10.3 highlights the four key metrics in the software delivery performance category that will help you quantify your team’s current performance and measure it against the industry benchmark:
Figure 10.3 – Four metrics to measure software delivery performance (source: Accelerate 2022 State of DevOps report)
It gives you an idea of how other companies are doing and if you’d like to aim higher by improving your current standings in any way.
However, all of this needs to be discussed and planned against the real needs of your application and the expectations of the end customers. So, context matters. Not all teams and applications require 10 deployments per day, but if they do, it’s necessary to make sure that your release processes support it.
The same report further describes a second clustering approach that accumulates the metrics from software delivery performance and sprinkles the reliability aspect on top of it. A holistic view of all these areas gives a representation of the current state of the workloads that you are managing, resulting in four different categorizations:
- Starting
- Flowing
- Slowing
- Retiring
As a DevOps practitioner, your goal should be to ensure that the result of this analysis matches the expectation from the workload or the value stream. These categorizations are highlighted in Figure 10.4:
Figure 10.4 – Clustering software delivery and operational performance metrics to identify the current state of the workloads (source: Accelerate 2022 State of DevOps report)