We summarized the discussion in five main takeaways for you.
1. SRE & DevOps Amplify Each Other
Site Reliability Engineering (SRE) is widely practiced, 52% of organizations that we surveyed were doing SRE to some degree or another. It’s also highly congruent and compatible with DevOps. Teams that practice SRE can achieve greater outcomes on those metrics studied in the DORA survey. There’s an independent movement between reliability work and software delivery work—they don’t necessarily predict each other but taken together they predict business outcomes.
Dave said: SRE is a force multiplier for software delivery and teams that have both of them really working together are 1.8 times more likely to report better business outcomes.
2. Shift Security Left
Incorporate security practices throughout the software development life cycle instead of trying to make it secure the day before the release. Last minute security checks can bring the entire launch to a halt and cause tension between teams. The better practice is to have security shifted left. In other words, do security all the way throughout the development process.
Shifting left doesn’t raise costs, in fact teams that shifted left successfully were 1.6 times more likely to exceed or meet those organizational goals and business outcomes.
How do you shift security left? Here are the best practices:
- Test for security
- Integrate security review into every phase
- Perform security reviews
- Build preapproved code
- Invite InfoSec early and often
3. What is Predictive of Performance?
Here are some of the things that are predictive of performance and optimize your DevOps practice:
- Loosely coupled architecture
- Trunk-based development
- Continuous testing
- Continuous integration
- Use of open source technologies
- Monitoring and observability practices
- Management of database changes
- Deployment automation
The State of DevOps Report research shows that organizations that undergo a DevOps transformation by adopting Continuous Delivery are more likely to have processes that are high quality, low risk and cost effective.
4. Internal Documentation Is Important Too
Internal documentation quality predicts a team’s success at implementing technical best practices.
Teams with quality documentation are:
- 3.8 times more likely to implement security best practices
- 3.4 times more likely to implement SRE best practices
- 2.4 times more likely to reach reliability targets
- 2.5 times more likely to fully leverage the cloud
One of the good things of COVID and how we shifted the way that we work is the focus on better documentation. Writing and communication plays a big role in software development.
KellyAnn said: Internal documentation is just as important as external documentation. Internal practices that actually help teams build better software and do it faster.
5. Don’t Neglect Culture
A generative organizational culture that optimizes for information flow, trust, innovation and risk-sharing is predictive of high performance.
The report found that high performing organizations are likely to encourage employees to take calculated risks without fear of negative consequences. Learning is the work and that idea has been reinforced by what we found in terms of outcomes now generative organizational culture.
Psychological safety is important. There is risk to moving fast and people want to feel safe and not feel like their job is always on the line. This is for everyone on the team.
People who feel included are half as likely to experience burnout during the pandemic.
KellyAnn said: You have to help people learn new things and treat them well enough that they’re not going to want to leave.
Dave said: Feeling follows action, right? And the same thing applies to the work; you do the action that is really consistent with the behavior and values and people start doing that. This is what we advise as a way to create that generative culture is to create tooling and processes that really reinforce those values.
One Takeaway From Our Guest Speakers
Tracy asked our speakers: “What would be your one key takeaway or parting thought for this discussion?”
Dave: Start small.
Find one bottleneck, find one capability, to improve and start with one team if needed. Try things, experiment, learn from it, and then share what you’ve learned throughout the organization. Share the successes and share the failures. It’s important to build that knowledge and share excitement around these ways of working, then it starts to propagate on its own.”
KellyAnn: Value documentation.
Value people who write it and give them credit. I very much hope that in the same way that the state of DevOps report made the idea that you can move fast with good quality code, that this is going to help give visibility to the: “oh my goodness if you have good internal documentation practices that can also help you move fast and have really good code.”
DevOps brings benefits to organizations and this continues to accelerate.
Everyone can be part of an elite team.
Elite teams are not random collections of elite individuals. Our research has shown that the highest performing teams are not when you draw the highest performing individuals and stick them together. It’s the team that’s elite and anyone can be part of an elite team. Any manager and any leader has the opportunity to nurture the things that will make their teams elite.