Key Findings from the 2021 DORA DevOps Report: SRE Practices, Documentation, Security, and Culture

This article is excerpted from the DORA DevOps Report 2021 (chapters 3‑4) and summarizes the survey’s key findings on technical operations, SRE, documentation, security, technical practices, and culture.

1. SRE and DevOps

When the DevOps community emerged publicly, Google was forming a parallel movement called Site Reliability Engineering (SRE). SRE shares many goals and techniques with DevOps and entered public awareness with the 2016 book Site Reliability Engineering . The study shows that SRE and DevOps are highly complementary, not conflicting, and together they improve team performance.

SRE emphasizes cross‑functional communication, psychological safety, and provides practical techniques such as Service Level Indicators/Objectives (SLI/SLO) and error‑budget‑driven prioritization.

Teams that adopt SRE practices are 1.4 × more likely to achieve higher SDO performance and 1.8 × more likely to deliver better business outcomes.

2. Documentation

High‑quality internal documentation (manuals, READMEs, code comments) is measured by three dimensions: helping readers achieve goals, being accurate/up‑to‑date/comprehensive, and being searchable, well‑organized, and clear.

Help readers achieve their goals

Accurate, current, and complete

Findable, well‑organized, and clear

Only about 25 % of respondents report high‑quality documentation, yet such teams are 2.4 × more likely to have better software delivery and technical operations (SDO) performance.

Improving documentation predicts higher success in security, reliability, SRE adoption, and cloud utilization.

3. Security

Security must be integrated throughout the software development lifecycle rather than added at the end. Elite teams are twice as likely to embed security practices and 1.6 × more likely to meet organizational goals when security is integrated.

Key security practices include automated security testing, integrating security reviews at every stage, building pre‑approved code libraries, and involving InfoSec early and often.

4. Technical DevOps Capabilities

The survey measured the following technical practices:

Loose‑coupling architecture

Trunk‑based development

Continuous testing

Continuous integration

Use of open‑source technologies

Monitoring and observability

Database change management

Deployment automation

Loose‑Coupling Architecture

Elite teams are three times more likely to use loose‑coupling, which reduces fine‑grained dependencies and improves resilience.

Continuous Testing & Continuous Integration

Continuous testing is a strong predictor of delivery success; elite teams are 3.7 × more likely to use it. Continuous integration increases the likelihood of meeting reliability goals by 5.8 ×.

Trunk‑Based Development

High‑performing organizations are 2.3 × more likely to adopt trunk‑based development, merging small batches of work daily.

Deployment Automation

Automation shortens delivery cycles, reduces deployment errors, and provides rapid feedback.

Database Change Management

Elite teams are 3.4 × more likely to manage database changes effectively, emphasizing collaboration and transparency.

Monitoring & Observability

Teams with strong observability practices are 4.1 × more likely to meet reliability goals and spend more time coding.

Open‑Source Technologies

Using open‑source tools makes knowledge transfer easier; elite teams are 2.4 × more likely to leverage them.

5. Culture

Culture is a primary driver of organizational and IT performance. A performance‑oriented culture, measured by the Westrum typology, predicts higher SDO performance.

Belonging and inclusion further enhance culture, reducing burnout and improving motivation.

Belonging & Inclusion

Investing in inclusive, belonging‑focused culture correlates with lower employee burnout and higher performance during DevOps transformations.