Change Failure Rate Explained

The third metric in the DORA metrics series is Change Failure Rate (CFR), which tracks errors or issues that cause problems for customers after a deployment to production. CFR is calculated as a ratio of the number of deployments that caused a failure to the total number of deployments. To improve CFR, teams can adopt measures such as improved testing, automation, infrastructure as code (IaC), deployment strategies like canary, blue-green, and rolling deployments, and feature flags. Automating testing and IaC is crucial for reducing errors and inconsistencies in the deployment process. Additionally, involving developers in production deployments helps them understand the connection between their changes and failures. Capturing information about outage duration and solutions used to restore service can provide valuable insights into root causes of failures and improve MTTR.