What We Learned Dissecting the World’s Most Popular Containers
summary
Analyzes the top 130 most popular Docker Hub images using container scanners including Docker Slim xray and common vuln scanners, modeling a CI/CD pipeline and timing scans. Finds a nearly perfect correlation between image size and scan time (~50 seconds per extra 500 MB, with scan time growing faster for very large images), some images exceed 2,000 known vulnerabilities, and about 20% of all found vulns are high/critical; many images are pulled billions of times.
AI-written summary · about
chapters (14)
- 0:00Intro
- 1:35Ayse Kaya Story
- 5:03About SlimAI
- 7:21Container Landscape
- 9:14Docker
- 10:58Container Management
- 13:14Methodology
- 17:12Container Size
- 19:29Container Complexity
- 20:27Box Plot Graph
- 21:52Package Landscape
- 23:30Attack Surface
- 24:36Vulnerability Severity
- 26:28What Comes Next
related talks
34mWhat We Learned Dissecting the World's Most Popular Containers
29mDay in the Life of a Base Image: The Evolution of Vulnerabilities in the Most Popular C...
52mThe Life of an Image: Evolution of Vulnerabilities in Popular Containers
35mSecuring Containers by Breaking In
47mSecuring Containers By Breaking In
49mVulnerability Exploitation In Docker Container Environments
29m5 Security Best Practices For Production Ready Containers
28mFresh Secrets From the Docks: Lessons Learnt From Analyzing 180,000 Public DockerHub Images