Practice Exams:

Microsoft AZ-104: Subscriptions, Policy, and Locks

Cloud & Architecture

Azure governance becomes easier to operate when subscriptions express accountability, Policy defines standards, and locks protect specific management operations. Confusing those controls leads to brittle environments and unexpected access errors. A subscription controls deployment and billing scope, a policy expresses desired guardrails, and a lock limits selected management operations. Combining them without understanding precedence can produce both accidental denials and false confidence about recovery or authorization.

On this page
  1. Begin with the boundary the business needs
  2. Use Policy to describe allowed configuration
  3. Distinguish assignments, initiatives and exemptions
  4. Treat locks as management-plane safety devices
  5. Keep RBAC and governance restrictions conceptually separate
  6. Build a subscription onboarding control sequence
  7. Review operational consequences before enforcing changes
  8. Rehearse one denied change and explain the result

Begin with the boundary the business needs

An Azure subscription is a management, quota, billing and access boundary. It is not simply a folder for resources. A company may split production and development, central networking and applications, or regulated workloads into different subscriptions because their operating requirements genuinely differ. Another company may use fewer subscriptions with clear resource-group ownership. The right choice depends on governance and service boundaries, not the number of teams on an organization chart. Subscription scope, Azure Policy and resource locks address different governance questions in Microsoft AZ-104; using one as a substitute for the others creates operational gaps.

Define what a subscription is responsible for: its business owner, environment, allowed regions, budget relationship, operational responders, and which platform policies it inherits. When every subscription has these characteristics, it is easier to interpret a deployment or cost alert. Without them, the estate may look tidy in the portal yet contain overlapping privileges and unmanaged workloads.

A subscription can belong under a management group, and the management hierarchy influences inherited policy and RBAC. Moving a subscription can therefore change effective controls. Review the new inherited rules and exemptions before moving it; do not assume the action is only a visual reorganization.

Use Policy to describe allowed configuration

Azure Policy evaluates resource configurations against definitions and assignments. It can audit, deny, modify or deploy related settings when supported by the particular effect and resource type. The purpose is to express enforceable or observable standards at the correct scope, not to replace permissions or a deployment pipeline.

Suppose the organization requires production storage accounts to use secure transport and approved network access. A Policy assignment can check the intended configuration and, where appropriate, prevent disallowed new deployments. It cannot by itself design the private DNS routing that an approved private endpoint needs. The workload team must still implement the service correctly and show evidence that policy compliance corresponds to the actual security objective.

Choose a scope that matches responsibility. A broad rule applied high in the hierarchy simplifies consistency but can affect services that cannot comply for legitimate reasons. An overly narrow assignment creates many ungoverned exceptions. Start by describing which resources should be covered, then design the assignment, parameters and exclusions to reflect that population.

Distinguish assignments, initiatives and exemptions

A policy definition expresses one compliance requirement. An initiative groups definitions that share a meaningful objective; its assignments can manage parameter values and effects together. Neither concept is a substitute for knowing what each individual definition does. The purpose of an initiative should be understandable to the people whose deployments it affects.

An exemption acknowledges a scoped, justified departure from an assignment without pretending the baseline requirement never existed. By contrast, an exclusion changes which resources are evaluated. Those choices have different audit and lifecycle implications. Record exception owner, reason, affected resource scope and review date; permanent exemptions with vague explanations undermine the governance system.

Deployment and remediation effects also need operational planning. Applying an assignment does not guarantee every existing resource instantly becomes compliant. Some resources require a remediation task, managed identity permissions or supported configuration changes. Verify what actually changed and whether the task introduced downtime before claiming that governance was corrected.

Treat locks as management-plane safety devices

Azure resource locks can block deletion or certain updates through the management plane, even when an RBAC role ordinarily permits the action. Microsoft distinguishes CanNotDelete and ReadOnly locks and explains how scope inheritance affects child resources. A delete lock helps guard a shared dependency against accidental removal, but a read-only lock should be tested against maintenance workflows because it can block legitimate configuration changes.

Locks are not data backups, immutability controls or an alternative to authorization. They do not automatically block every data-plane operation on the protected service. A storage account under a management lock can still require blob versioning, retention and appropriate data access. Calling a resource lock “ransomware protection” without examining the data path is misleading.

Inherited locks deserve attention. A group-level lock may protect an important database while unintentionally preventing temporary test resources from being deleted. Some extension resources inherit parent restrictions too. Before locking a broad group, list the real maintenance tasks performed underneath it and test which Azure API operations those tasks call. A lock should create intentional friction, not unbounded operational paralysis.

Keep RBAC and governance restrictions conceptually separate

Azure RBAC grants actions to identities at scopes. Policy evaluates configuration and sometimes enforces standards, while locks restrict particular management modifications. One control cannot be read as a complete replacement for the other two. An operator might have Contributor but still receive a policy denial; another might have sufficient role permissions to update a resource but be stopped by a ReadOnly lock.

When troubleshooting a failure, determine which layer rejected the request. A failed deployment response may name a policy assignment or effect; a role failure may indicate insufficient Azure RBAC permission; a lock conflict typically identifies the protected scope. Granting broader Owner privileges does not solve every policy or lock denial and may introduce unnecessary risk.

Use Azure RBAC design to keep access narrow. Reserve governance authority for people responsible for the standard itself, not every application administrator. Teams need a supported route to request an exception without turning policy changes into ad hoc high-privilege portal edits.

Build a subscription onboarding control sequence

A new production subscription should enter the intended management group, inherit baseline policy, receive deliberately scoped administrative groups, connect to shared network and monitoring capabilities if needed, and have a named business cost owner. These steps are easier to make repeatable through approved automation than through a sequence of remembered portal clicks.

Test the sequence with one harmless deployment that should pass and one deliberately noncompliant resource that should be denied in the nonproduction test environment. Review policy compliance, effective role assignments and any locks. Ensure that the deployment team can see enough error detail to understand the restriction without receiving broad access to override it.

Lifecycle also matters. If a workload is retired, decide what must be retained, which role assignments should end, which shared networking links must be removed, and which locks require authorized removal before cleanup. A lock that successfully prevents accidental deletion can become a migration obstacle when no one knows who owns it.

Review operational consequences before enforcing changes

A proposed policy or lock can be technically correct but operationally disruptive. For example, enforcing a network restriction across existing storage accounts without checking dependent automation can break backups or deployments. Apply changes through review, staged testing and measurable rollout, with a supported rollback and documented exception path where necessary.

A governance dashboard may show high compliance while application owners still experience repeated failed changes. Examine those failures and distinguish valid restrictions from overly broad assignments. Conversely, a large list of exemptions may signal that the baseline is unrealistic or that the onboarding process is inadequate. Governance health includes the quality of the operating experience as well as raw compliance percentages.

Review scope periodically as teams and workloads evolve. Subscriptions may acquire new owners, regions may gain or lose approved status, and product configurations may change. The governance model should retain a stable vocabulary while permitting controlled updates to actual controls.

Rehearse one denied change and explain the result

In a disposable Azure resource group, grant a narrowly scoped role, assign an appropriate audit or deny policy for a test configuration, and apply a delete lock to a resource selected for the exercise. Attempt a change that violates each control separately. Observe which system reports the denial and which permissions are required to change the governing rule itself. Remove the test controls safely afterward.

A subscription defines an operating boundary; Policy evaluates configuration against rules, RBAC grants principals permissions, and locks constrain selected management operations. Test each control against a concrete proposed change rather than assuming their effects are interchangeable. Clear ownership, narrow access and a documented exception process keep governance useful when the estate expands or an emergency change is required.

Continue learning

Related guides

Azure Resource Locks and Operational SafetyUse Azure resource locks without blocking authorized recovery.Azure Policy Initiatives for Platform TeamsBundle Azure Policy controls with exclusions and remediation.Azure Management Groups at ScaleApply inherited governance across Azure management groups and subscriptions.

Related Posts

• Legacy MCSA Cloud Platform: Azure Architecture Skills and Current Paths

• Top Cloud Certifications Valuable in 2022: Is Microsoft a Leader?

• Microsoft AZ-104: What Azure Administrators Actually Do

• Microsoft AZ-104: Is Azure Administration the Right Career?

• Conquer the Cloud: Your Beginner’s Blueprint for Microsoft Azure Mastery

• Microsoft AZ-104: Six-Week Study Plan

• Azure Mastery Beyond the Basics: Governance, Green Cloud, and Strategic Growth

• Cloud-Native App Architecture Using Microsoft Azure Cosmos DB

• Cloud Strategy Unlocked: How AZ-305 Shapes Real Azure Architects

• Cloud Confidence Starts Here: A Four-Part AZ-900 Certification Blueprint