> I know why orgs install central management systems (ease of maintenance at scale). But when one system can be used to compromise all the nodes in an org, is i

> I know why orgs install central management systems (ease of maintenance at scale). But when one system can be used to compromise all the nodes in an org, is it really worth it? If you think about it, the modern cloud-native approaches are logically aiming for the same. You have a centralised management system (the cloud provider's engine), with programmatic configuration (infrastructure-as-code and indeed everything-as-code). We even call these modern best practices. Just instead of a webgui management console, we have git repos, and in place of two-person change protocol we have mandatory code reviews enforced by tooling. A single point of failure is one side of the coin. The side you're not looking at is single point of control.