The State of Infrastructure Drift: What We Learned From 100+ DevOps Teams
summary
Survey of 150 teams and interviews with 50+ teams finds drift common: nearly half see it weekly and 41% once a month, a third lose ~1 hour every week or two, and ~20% of incidents are security issues. Concrete failures include unnoticed IAM access keys and policy attachments, security groups opened in the console, and S3 versioning turned on that spikes bills. The project driftctl compares Terraform state to cloud APIs and reports coverage drops (example 100%→75%).
AI-written summary · about
chapters (13)
- 0:00Intro
- 1:37Why is Drift Happening
- 2:56Who is Stefan
- 3:27What is Infrastructure Drift
- 3:51Main causes of Infrastructure Drift
- 5:47How often it happened
- 7:41Solutions
- 10:18Stories
- 10:48First story
- 16:07Second story
- 20:05Third story
- 24:14Drift Control
- 26:18Outro
related talks
31mWhy You Should Take Care of Infrastructure Drift
24mWhy you should take care of infrastructure drift
30mWhy you should take care of infrastructure drift
30mDon't Panic: Getting Your Infrastructure Drift Under Control
21mFrom Ghost Assets to Infrastructure Drift - Don't Get Spooked
18mScaling DevOps in Complex Environments: Challenges, Failures, and Successes
42mLessons Learned as a DevOps Consultant
29mManaging infrastructure with Git: First principles of GitOps