Policy Evaluation Logic
What you'll learn: How cloud providers decide whether to allow or deny your requests using a specific decision-making process.
When you try to access a cloud resource—like reading a file from storage or starting a virtual machine—the cloud provider's IAM system must answer one question: "Is this allowed?" It doesn't just check one policy and call it done. Instead, it follows a strict policy evaluation logic to make that decision.
The Three-Step Decision Process
Think of IAM like a security checkpoint with three gates:
-
Explicit Deny (Gate 1): First, the system checks all policies for an explicit
Deny. If any policy explicitly denies the action, the request is immediately blocked—no exceptions. This always wins. -
Explicit Allow (Gate 2): If no explicit deny exists, the system looks for an explicit
Allow. If at least one policy grants permission, the request can proceed. -
Default Deny (Gate 3): If there's no explicit deny and no explicit allow, the request is denied by default. This is the "fail-safe" position—you must be explicitly granted permission.
Why This Order Matters
Explicit denies acting as a trump card means you can broadly grant permissions, then carve out specific exceptions. For example, you might allow a user to access all S3 buckets except the finance bucket—the explicit deny on finance overrides the broader allow.
Key Takeaway: Cloud IAM uses a three-tier evaluation: explicit deny always wins, explicit allow comes next, and if neither exists, the default deny blocks the request. This "deny-by-default" approach ensures security unless permissions are purposefully granted.