Skip to main content

Frequently Asked Questions

How does GCP Organization Policy differ from IAM if both can restrict what happens in a project?

IAM answers 'who can perform this action' by granting roles to identities, while Organization Policy answers 'what configurations are allowed to exist at all,' regardless of who's making the request — even an Owner-level IAM identity can't create a resource that violates an active Organization Policy constraint, like `constraints/compute.vmExternalIpAccess` blocking external IPs. This makes Org Policy the layer for guardrails that must hold even against a misconfigured or overly permissive IAM grant.

What's a common mistake teams make when applying Organization Policy constraints?

Applying a restrictive constraint at the Organization root without first checking which existing projects would break is a frequent cause of outages — a policy blocking external IPs, for instance, can immediately strand VMs that legitimately need public access, like a bastion host or a public-facing service. Testing constraints at a Folder level against a subset of projects, and using dry-run/audit mode where available, is the safer rollout path before enforcing organization-wide.