Last year, I did some hard things. Some really uncomfortable, hard, things. One of them was to present at the Bell Network and Technology Leadership Conference. I am not very comfortable with public speaking (is anyone really?), however, I have wanted to conquer this fear by attacking it head-on. Not only did I jump into this to tackle this fear but this was an intentional goal I set years ago. I feel extremely grateful to have had this opportunity (thank you so much if you were there!).
Standing in front of 100 technology leaders and talking about my passion for DevOps and, specifically, the Role of DevOps in Cloud Transformation, was somewhat of a dream come true — to have the experience and confidence to stand in front of interested parties and share my passion.
Speaking to a group without prior knowledge of their level of DevOps maturity, I thought it best to level set on what my definition of DevOps was before jumping into how to use DevOps practices to achieve objectives and how to measure success.
My presentation, titled “The Role of DevOps in Cloud Transformation”, had three main takeaways:
- What is DevOps
- How DevOps helps achieve business and IT objectives
- How to measure the value of DevOps investments over time
Overall, the presentation was well received. I felt a reciprocal energy from the audience that led me to believe what I was saying resonated with them. In addition, there were many thoughtful and forward-thinking questions asked after the presentation. I want to share three of these questions with you today, as I think they’re a wonderful representation of different facets of DevOps implementation that are elusive and thus difficult to grasp.
Question 1: I am being told to leverage DevOps in my Operations Teams. How can I introduce DevOps practices into our operational processes? How are DevOps practices going to address our challenges? Not only now, when we’re operating mostly in our data center, but as the applications we are managing are being migrated into the Public Cloud?
In many organizations, development, infrastructure, and operations teams are operating in separate silos. This is still a very prevalent organizational structure in large enterprises, especially ones that have been around 100 years! (Bell Canada was founded on April 29, 1880 – 142 years ago!).
For this particular question, the Operations lead felt there was something DevOps had to offer his team but wasn’t sure what. To answer the question, I began with the software development lifecycle and pointed out how DevOps practices can be applied at every stage, there are no constraints on where they can or should be applied – DevOps best practices can and should be considered at every phase of the software development life cycle.
What you see on the far right is ‘Operate’. In order to successfully operate software applications, the principles of DevOps (flow, feedback, and continuous learning) still apply, although a lot of the time, this is the neglected part of the lifecycle.
Keeping all three DevOps principles in mind, this is how I answered this question:
Pertaining to flow
Are there defined operational requirements for the development teams that contribute to data being available for debugging? Do you have meaningful, structured log data? Do you have a correlation between traces, logs, alarms, and the end-use application that is being impacted? Additional, more modern, observabilities tooling might be required or additional integration points.
Pertaining to feedback
Can your team decipher what the impact is on the end user and the business impact (and if there is a financial impact)? If not, then there is work to do to ensure this flow of data is enabled. How to do this depends on the existing tooling and skill set of the team.
Pertaining to continuous learning
Also, there may need to be closer collaboration between the dev and ops teams with the goal to define and work through requirements that the ops team mandates from the dev teams.
“Even though many people see DevOps as fundamentally addressing technological aspects of automation and tooling, only organizations that also address fundamental misalignments between teams are able to achieve the full potential benefits from adopting DevOps.”
— p .66 – Team Topologies
The DevOps capabilities that will drive improvement to your Operations functions are:
- Gathering and implementing customer feedback
- Having a lightweight change approval process
- Monitoring across applications & infrastructure to inform business decisions
- Check system health proactively
- Monitor quality and communicate with the entire value stream team continuously and make it visible
- Support a generative culture (high cooperation, good information flow)
Question 2: How can I increase adoption by the product teams (application development teams) of the DevOps CI/CD pipelines my team has built?
There is a DevOps team in your organization. They have defined and implemented a CI/CD pipeline and processes in place for onboarding developers to the pipeline. Being proud of their work and a shiny new pipeline, they’ve offered it up to the developer teams in hopes folks will jump on the bandwagon. Maybe some did, for some short period of time, then usage dropped off to near zero and teams resumed using their shadow IT processes. Why?
This is an all too common challenge facing many organizations that have jumped head first into DevOps implementation by forming a DevOps Team by forming a separate DevOps team to ‘serve’ the organizational infrastructure and DevOps requirements. Where enterprises are siloed and functions are centralized, which is an anti-pattern for DevOps enablement, we run the risk of worsening disconnection and disruption of flow. (Anti-patterns for DevOps enablement were defined by Matthew Skelton and Manuel Pais originally documented on this website).
The “DevOps team” anti-pattern Type B can easily become a bottleneck or increased source of disconnect.
There are instances where a DevOps silo can work. When this team is mandated with clear objectives to bring Dev and Ops closer together, helping to bridge gaps with process requirements, automation, and tooling.
Question 3: How can I use DevOps practices to make our testing cycles more effective and efficient?
Automated testing and orchestration is a critical success factor in delivering high-quality applications and achieving high levels of customer satisfaction, which in turn builds trust and minimizes churn. However, this is often times treated as an afterthought. Luckily, the foundational DevOps principle of feedback offers guidance when it comes to how to establish greater levels of quality. Enabling fast and reliable automated testing as part of your CI and CD pipelines is the goal. Shifting testing left is best, making sure to test code as close as possible to development. This creates a fast feedback loop that helps developers find and fix errors early when there are the fewest constraints (time and resources). Ideally, this is done as part of the CI pipeline (CI+) where automated test suites are executed every time code changes are checked into the main, which triggers a deployment to a test environment that mimics production where integration test suites are executed.
In summary, the DevOps capabilities that will drive improvement to your testing/QA functions are:
- Implement test automation
- Support test data management
- Implement Continuous Integration with test automation
- Implement Continuous Delivery into test environment mimicking production
- Work in small batches
- Support a generative culture (support learning and conscious inquiry, trust, and cooperation)
I have to say, this was the best conference and tradeshow I’ve ever attended. Congratulations to Bell for investing in their team and their sponsors and for providing outstanding leadership, vision and transparency.
It was a pleasure to share my passion for technology in business with everyone in attendance. I met wonderfully talented people who truly care about their colleagues and the overall vision of being “Better Together”.
At Indellient, we invest heavily in our DevOps practice to create foundational cloud transformation for our clients. We believe the outcomes of DevOps: velocity, resilience, automation, compliance, and rapid changes contribute to a competitive advantage, which is why we architect and implement DevOps culture and best practices that leap our partners forward and enable a cloud operating model.
2630 Bristol Cir #100 Oakville, On L6H 6Z7
+