I feel like almost all issues in this article stem from not following ops best practices. And similar issues would happen no matter what technology was being used, they just happen to blame their issues on docker because that's the tool they were using at the time.