Some today claim that DevOps is dead. Emily Freeman isn’t buying it. If you’re a cynic, you may think: well, yeah. Of course not. She literally wrote the book on it! But before jumping to conclusions, you’d do well to understand her argument.

Is DevOps dead?

Freeman’s certain that those who make this claim do so because they work with newer and smaller companies, companies who naturally pick up on the best practices hard won over the last ten years. However, from her consulting work with larger and older organizations, she knows that many teams are still far behind.

And for the skeptics out there, she’s got the data to prove it: 56% of organizations are still in the medium (44%) to low (12%) performers category, according to Google’s 2019 Accelerate State of DevOps Report.

She also wonders why there were 81 DevOpsDays events just this year. Which does seem like a lot for something that’s dead.

Rather than being dead, Freeman is convinced something else is afoot. Yes, most of us understand the benefits of both DevOps and agile these days. But the reality is while we get it in theory, most of us are still far from perfecting it in practice.

This is the state of things in 2019, and it’s here where Freeman wants to intervene. During her keynote address at Sitecore Symposium 2019, she not only laid out where we are, but also gave us a clear path forward.

An agile reminder

The first thing she reminded us of are the 12 principles of agile (if you’re well versed, feel free to skip ahead):

  1. Satisfy the customer by continuously delivering beneficial software

  2. Accept and welcome changing requirements across the process

  3. Deliver working software frequently

  4. Enable developers and business stakeholders to work in daily cooperation

  5. Trust your engineers to get their work done

  6. Convey information face to face (or via camera)

  7. Realize that working software is the most important measure of success and progress

  8. Maintain a constant pace of work

  9. Strive for technical excellence

  10. Simplify requirements and features

  11. Allow teams to self-organize for the best product

  12. Regularly reflect, as a team, on how to become more effective and adjust behavior accordingly

The point of all of the above was to breakdown silos, creating fluidity and efficiency between teams. The problem is that tech changes exponentially, while people and cultures don’t. This is one of the reasons why Freeman is quick to remind everyone she speaks to that DevOps is about people first, processes second, and technology last.

A philosophy to speed development

She’s also quick to remind us that DevOps is a philosophy and methodology — not a prescription. It’s a method for creating a process that brings everyone to the table and transforming the way they work together. The goal is a non-linear, fluid, and iterative process that creates better software, faster and (usually) cheaper.

Ideally, teams work in concert while adapting to the ever-changing needs of the business and the market. DevOps began, at least in part, as a way for development teams to work with operations teams better. For Freeman, it can’t be dead if it still hasn’t arrived. While we have yet to reach the DevOps Promised Land, we have learned a lot in the past ten years since it came on the scene.

Freeman outlined several key principles that everyone — from the novice to the expert — can learn from. Freeman offered snapshots from three years of the Annual State of DevOps reports put out by Google.

In 2014, the key things that led to high performance were:

  • Peer-review change approval process

  • Version control for all production artifacts

  • Proactive monitoring

  • High-trust organizational culture

  • Win-win relationship between dev and ops

The 2017 key findings include:

  • Transformational leaders shape an organization’s culture and practices

  • High-performance teams achieve both faster throughput and better stability

  • Automation is a boon to organizations

  • Loosely coupled architectures assist continuous delivery

  • Lean product management drives higher organizational performance

In 2019, elite performers deployed multiple times per day, had a short lead time, were able to recover from failure in less than an hour, and had a change failure rate of 0-15%.

However, lead time has increased year-over-year lately. Why? One theory is that this work is hard. From when a developer commits code to when it’s actually deployed in production.

 

 

The 2019 key findings:

  1. The industry continues to improve among elite performers

  2. Delivering software quickly, reliably and safely at the pace of technology transformation and organizational performance

  3. The best strategies for scaling DevOps in organizations focus on structural solutions that build community

  4. Cloud continues to be a differentiator for elite performers and drives high performance

  5. Productivity can drive improvements in work/life balance and reductions in burnout, and organizations can make smart investments to support it

  6. There’s a right way to handle the change approval process, and it leads to improvements in speed and stability and reductions in burnout

Number two is the heart of DevOps. Delivering software quickly and agility. This is where continuous delivery and continuous development (CD/CI) comes in. According to Emily, CI/CD includes thorough and automated testing in a robust test suite. The benefits are clear: an accelerated feedback loop, a decreased interpersonal conflict, and a reliable deployment process.

Is it too late to get started? If not, how?

If 43% are in the high to elite performance this year, then some may wonder if it’s too late to get started. According to Freeman, no. Even the elite performers aren’t perfect. And the vast majority of us are still struggling in most areas, if not in all.

In addition to the key findings of 2019, she wanted to stress four points:

  1. Run tests

  2. Invest in training engineers to be agile

  3. Automate releases

  4. Write documentation

These are important, but Freeman stressed, again, that DevOps is about people before process and process before technology. On this front, she offered great advice for diagnosing culture.

One of the greatest challenges business face

This is one of the greatest challenges companies face: identifying the actual state of their culture. Freeman suggests looking at yours with a cynic’s eye. Really discovering your culture requires asking your employees what they really think. Anonymous surveys are best. And if you expect people to actually take them, only add relevant, smart questions. Questions like, “How likely would you be to leave for a 10% raise?” can be incredibly revealing.

She also suggests observing interpersonal communication — are people engaged, respectful, assuming positive intent from one another? And looking at leadership. After all, culture flows from the top.

Getting culture right

But diagnosing culture is only one half of the equation. What about fixing it? Which just begs the next question: what does a healthy culture look like?

She offered four principles as guidelines (which I think every employee would get behind):

  1. Demand diversity
    Of age, experience, gender, race, etc. In addition to decreasing unconscious bias, it offers many other business advantages.

  2. Be reasonable — please!
    Your people shouldn’t work more than 40 hours a week. Off-hours work and deployments are unnecessary — at least the vast majority of the time. They’re a symptom of larger problems in your organization.

  3. Provide great benefits.
    Medical insurance matters. Engineers are disproportionally affected by mental health issues like depression and anxiety.

  4. Encourage alternative thought
    This ties back into number one. For example, junior Engineers bring a new perspective, often asking questions about things seasoned engineers never thought of.

For those who are concerned with the business value, Freeman offers this: happy engineers make great software.

What’s next?

In closing, Freeman asked what’s next. While she was clear that she’s no futurist, she does think that what’s next is reliability, specifically site reliability engineering (SRE). She thinks that in the next few years, SRE will be as important of job title as DevOps is today. This has to do with everything from your site not crashing to your data being consistent.

Overall, it was a great opening to the technical track, and I’m looking forward to the week ahead.

Mohan Kasibhatla is Vice President of Product Marketing at Sitecore. Find him on LinkedIn.