100% Latest & Updated Microsoft AZ-104 Practice Test Questions, Exam Dumps & Verified Answers!
30 Days Free Updates, Instant Download!
Question 60
An established Microsoft Entra production tenant has a verified custom domain and is not in the first 30 days of a trial. Ordinary users require only one authentication method for self-service password reset. A Global Administrator with two registered methods cannot use that ordinary-user setting to reduce the administrator reset to one verification. Which policy boundary explains the difference?
- Add the Global Administrator to an ordinary-user SSPR-enabled group whose policy requires only one method, expecting group membership to override privileged-role requirements
- Accounts assigned specified administrator roles follow a separate stronger SSPR reset policy that the ordinary-user setting does not relax
- Disable a method in the tenant Authentication methods policy so the administrator SSPR process automatically drops from two verifications to one
- A Global Administrator must always present exactly three verification methods during SSPR, regardless of tenant configuration
Explanation
Correct Answer
B:Microsoft Entra applies a separate, stronger administrator SSPR reset policy to designated directory roles such as Global Administrator. In an established tenant with a verified custom domain, the administrator flow requires two supported authentication data gates, not the ordinary one-method user setting. Exceptional one-gate circumstances for new trials or default-domain tenants have been excluded explicitly by the scenario.
Incorrect Answers
A:Scoping the administrator into an ordinary-user SSPR group does not remove the separate administrator reset policy. In this established tenant the administrator must satisfy the privileged two-gate reset checks even when ordinary users are allowed a one-method flow. Group inclusion can govern whether users are enabled for a standard user policy, not rewrite the administrator-specific verification standard.
C:The administrator-specific SSPR policy does not depend on the tenant Authentication methods policy in the same way as ordinary user method availability. Disabling a method there does not change the required two separate authentication data gates for this Global Administrator. It might affect other sign-in or registration experiences instead, without solving the requested privileged reset-policy restriction.
D:The stronger standard administrator reset policy generally requires two pieces of authentication data, not an unconditional three. The question specifies a verified-domain established tenant, so the normal two-gate rule rather than a new-trial one-gate exception applies. Requiring exactly three would invent a threshold that this Microsoft Entra SSPR scenario does not support.
Question 61
A Microsoft Entra tenant has completed migration from the separate legacy MFA and SSPR authentication-method settings to the unified Authentication methods policy. The identity team wants to enable SMS as an allowed SSPR proof only for one approved pilot group; the SSPR Properties setting already controls which users are eligible for reset. What should the team configure?
- Change the SSPR Properties setting to All users and rely on the SMS method being familiar to the pilot
- Enable the SMS method for the pilot group in the unified Microsoft Entra Authentication methods policy
- Edit the legacy SSPR Authentication methods list instead, since those old method switches still govern a migrated tenant
- Increase the number of methods required to reset from one to two without permitting SMS in the unified policy
Explanation
Correct Answer
B:Once migration is complete, the unified Authentication methods policy controls allowed methods for the targeted groups across supported authentication flows, including SSPR. Enable SMS for the specific pilot group there while leaving the already-established SSPR eligibility scope unchanged. Users still have to register SMS and satisfy the number of required recovery challenges.
Incorrect Answers
A:Setting SSPR Properties to All would change the separate eligibility boundary and could include employees outside the intended reset population. It does not supply group-scoped SMS method authorization in a tenant using the unified policy. Eligibility for SSPR and which verification methods are enabled are distinct policy decisions and must both be correct.
C:Migration Complete means the legacy MFA/SSPR authentication-method controls no longer determine ordinary method availability for the migrated tenant. Editing those settings cannot establish the requested group-scoped SMS authorization. The team needs the unified Authentication methods policy rather than an obsolete method switch; exceptions such as legacy security questions are separate.
D:Changing the required count alters how many approved proofs a user must provide, not which methods are made available. If SMS is disabled for the pilot in the unified Authentication methods policy, increasing the count will not make it usable and could further reduce employees' ability to complete SSPR.
Question 62
A help desk is asked to provide Microsoft Entra self-service password reset to employees whose accounts currently exist only in on-premises Active Directory, with no corresponding Entra user objects. The organization wants subsequent supported resets to write back to AD DS. What must be established before password writeback can serve those employees?
- Enable password writeback alone and assume the feature creates a new Entra user for each on-premises account
- Turn on password hash synchronization but explicitly exclude all affected users from directory synchronization and SSPR enrollment
- Synchronize or otherwise establish eligible Entra user identities, configure SSPR scope and methods, and then integrate supported password writeback
- Add the user names to the Selected SSPR group only in local Active Directory without synchronizing any matching cloud identities
Explanation
Correct Answer
C:The employees first need corresponding eligible Entra user identities, a supported SSPR configuration, and registered verification methods. A supported Entra Connect or cloud-sync integration must then implement writeback so cloud password-reset changes reach the original AD DS accounts. This sequence addresses identity presence, secure recovery and hybrid password consistency without treating writeback as provisioning.
Incorrect Answers
A:Writeback is a password-change delivery mechanism from a supported cloud-recognized identity to an on-premises directory; it does not create Microsoft Entra user objects for accounts that exist only in AD DS. Without corresponding identities eligible for cloud SSPR, there is no supported reset event from which the writeback service can propagate a password.
B:Password hash synchronization primarily supports the on-premises-to-cloud credential path. Explicitly excluding the employees from directory synchronization leaves the cloud without their identities, and omitting SSPR enrollment leaves them unable to complete its recovery workflow. Enabling a password-related sync feature alone does not create eligibility or reverse reset changes.
D:A group membership that exists only in an unsynchronized on-premises directory is not automatically an eligible Entra SSPR assignment for an absent cloud user. The Selected option applies to Microsoft Entra groups and recognizable tenant users. Supported identity synchronization and enrollment must precede a workable hybrid SSPR reset and writeback flow.
Question 63
A platform engineer must inspect the configuration and running state of all virtual machines in one resource group. The engineer must not start, stop, modify, or delegate access to any resource. Which Azure built-in role should you assign at the resource-group scope?
- Virtual Machine Contributor
- Reader
- Contributor
- Role Based Access Control Administrator
Explanation
Correct Answer
B:Azure Reader permits control-plane inspection of resource properties and running state at the assigned resource-group scope, including VMs within the group. It does not grant write, delete, VM start/stop or Azure RBAC role-assignment privileges. This matches the engineer’s requirement to examine every VM in one resource group while leaving administration and delegation with other personnel.
Incorrect Answers
A:Virtual Machine Contributor permits virtual-machine management operations such as start, stop and configuration changes at its assigned scope. Although that role can inspect a VM, it would also let the platform engineer change power state or related VM settings, violating the explicit read-only requirement. Choosing a workload-specific Contributor role is therefore not least privilege when the task is inspection without operation.
C:Contributor is broader than required: it allows most Azure resource management operations across the assigned resource group, including VM modifications and deletion, even though it excludes ordinary IAM role assignment. A platform engineer with Contributor could change the machines being inspected rather than simply view them. The read-only Reader role achieves the necessary visibility without authorizing disruptive changes.
D:Role Based Access Control Administrator manages Azure role assignments and related authorization operations. It is appropriate for an access officer who must delegate roles but does not need general VM administration. This engineer explicitly must not delegate access and only needs to inspect VM state, so a role-management role would provide the wrong permissions while failing to improve the read-only inspection task.
Question 64
An operations team manages existing Azure resources using Contributor. Its access officer must independently grant and revoke resource-level role assignments but must not receive permissions to deploy or delete those resources. Which built-in Azure role should the access officer receive?
- Contributor
- Virtual Machine Contributor
- Role Based Access Control Administrator
- Reader
Explanation
Correct Answer
C:Role Based Access Control Administrator grants the role-assignment management capability needed to create and remove Azure RBAC grants at the specified scope, without the broad resource-deployment operations of Contributor or Owner. Assign it at the smallest resource-group or resource scope that covers the access officer’s delegated responsibilities, and consider role-assignment conditions when further restricting permitted roles or principals.
Incorrect Answers
A:Contributor grants broad resource management Actions but does not normally include Microsoft.Authorization/roleAssignments/write, so it is not the standard role for creating or revoking Azure RBAC assignments. Assigning Contributor to the access officer would permit resource changes the officer must not perform while still omitting the IAM delegation capability that the scenario requires.
B:Virtual Machine Contributor supports VM deployment and operational management within its scope, including power and configuration actions. It is not an IAM authorization administration role. The access officer is responsible for granting and revoking resource-level role assignments, not for administering VM state; providing VM Contributor would be both irrelevant to that duty and unnecessarily permissive.
D:Reader provides visibility into Azure resource metadata and existing access configurations but cannot create or delete role assignments. It is valuable for audit personnel who inspect authorization, not for officers who must independently grant and revoke access. The requested IAM writes require a role containing the appropriate Microsoft.Authorization operations, with a scope appropriate to the affected resources.
Question 65
A Bicep deployment creates a user-assigned managed identity and immediately gives that identity an Azure RBAC role at the resource-group scope. Some runs return PrincipalNotFound even though the principalId is correct and the deployment account already has role-assignment write rights. What role-assignment configuration addresses this documented directory-replication timing problem without increasing permissions?
- Temporarily grant Owner to the deployment account at subscription scope, even though role-assignment write permission already exists
- Place the managed identity application ID rather than its principal object ID in the role-assignment principalId field
- Set the role-assignment principalType property to ServicePrincipal with a supported API version such as 2022-04-01
- Add dependsOn for the identity resource and assume resource deployment ordering also guarantees cross-region Entra principal replication
Explanation
Correct Answer
C:Microsoft documents that a newly created service principal or managed identity can trigger PrincipalNotFound before directory replication completes. Specifying principalType as ServicePrincipal in the role-assignment resource using a supported API version lets ARM handle that known timing scenario more reliably while retaining the same principal ID, assigned role and least-privilege scope.
Incorrect Answers
A:The deployment account already has Microsoft.Authorization/roleAssignments/write permission, so broadening it to subscription Owner would not resolve a transient PrincipalNotFound response. It introduces excessive management and delegation privileges while the actual failure is directory replication timing for the newly created identity principal.
B:The principalId property identifies the Entra service principal object that receives the RBAC role, not an application/client ID or the friendly managed-identity resource name. Substituting another identifier can point authorization at the wrong principal or fail validation; it does not address cross-region propagation of the new object.
D:A deployment dependency orders creation of the managed identity before the RBAC extension resource, but it does not guarantee that the backing Entra service principal has replicated to every authorization region at that instant. The documented mitigation is to specify the new principal type, not to assume deployment ordering is a directory-replication barrier.
Question 66
An analyst can list an Azure Key Vault resource in the portal using an Azure Reader role but receives a forbidden response when requesting a secret value. The vault uses Azure RBAC for its data plane. What distinction explains the failure?
- Azure Reader at subscription scope automatically confers Key Vault secret-value read permission on all vaults in that scope
- A Key Vault Reader data role includes the ability to retrieve the contents of secrets, not merely their metadata
- Azure Reader permits management-resource inspection but not the vault secret data action
- The Key Vault RBAC permission model merges control-plane Reader access into data-plane secret retrieval by default
Explanation
Correct Answer
C:The vault uses the Azure RBAC data-plane permission model, so the analyst must have a compatible role authorizing secret retrieval separately from Azure Reader. Reader accounts for the ability to list the vault resource in the portal; it does not supply the data-plane permission for the forbidden secret-value operation. This distinction applies even to the same named vault.
Incorrect Answers
A:Azure Reader is a control-plane role for viewing resource configuration and other management metadata. Giving it a broader subscription scope would let the analyst see more vault resources but would not add the data-plane permission needed to call Get Secret and retrieve a secret value. Scope expansion cannot manufacture missing data actions.
B:Key Vault Reader is a vault data role that reads metadata for keys, secrets and certificates but explicitly cannot read sensitive secret values. Confusing metadata visibility with Get Secret is the same access-boundary error observed in the case. A Key Vault Secrets User permission is an example of a role granting secret-value reads when assigned appropriately.
D:The management plane and vault object/data plane have separate authorizations. Selecting Azure RBAC as the vault authorization model does not merge their role sets or convert Reader into a secrets reader. Without a relevant vault data role, the analyst can continue to inspect the vault resource while secret-value access remains forbidden.
Question 67
A subscription Owner can normally delete resources. In one deployment stack, deleting a managed resource fails and IAM lists an inherited deny assignment for resource deletion. The owner wants to understand the failure before changing permissions. Which conclusion is correct?
- An Owner role always overrides a deny assignment when the action is at subscription scope
- The delete must succeed if the user also has User Access Administrator
- The deny assignment can block the operation despite the inherited Owner grant
- Change the resource group CanNotDelete lock because an inherited deny assignment must be only a lock
Explanation
Correct Answer
C:Azure evaluates both allows and denials; an applicable deny assignment wins even when a role grants the delete action. Inspect its scope and protected-resource configuration instead of broadening user access.
Incorrect Answers
A:Owner provides broad role grants but does not override an applicable Azure RBAC deny assignment. Repeating the action with the same permissions would leave the prohibition in force.
B:User Access Administrator governs access management, not an unconditional ability to bypass protection imposed by deny assignments. Adding another role cannot cancel that denial.
D:Management locks and deny assignments both can prevent deletion, but the scenario explicitly identifies a deny assignment. Removing an unrelated lock would not erase the applicable RBAC denial.
Question 68
A network operations contractor must modify NSGs, routes, and VNets in one resource group but must not grant IAM roles or modify the virtual machines there. What Azure built-in role best matches those permissions?
- Contributor at subscription scope
- Network Contributor scoped to the networking resource group
- Role Based Access Control Administrator at that resource group
- Reader at the networking resource group, assuming visibility into NSGs permits modifying their rules
Explanation
Correct Answer
B:Network Contributor grants Microsoft.Network management actions without role-assignment management and without general VM management. Resource-group scope confines those network changes to the assigned group.
Incorrect Answers
A:Contributor allows broad management of many resource providers across the subscription, including compute. The contractor needs the network provider actions within one resource group, not broad deployment privileges.
C:This role focuses on role assignments and does not supply the required broad Microsoft.Network management actions. It would enable the very access delegation that is prohibited.
D:Reader lets the contractor inspect NSGs, route tables and VNet properties but lacks the management write Actions needed to modify them. Network Contributor at the targeted resource-group scope provides the relevant network writes without general compute or IAM delegation.
Question 69
An automation identity holds a custom Azure role where Actions includes Microsoft.Compute/*, while NotActions removes Microsoft.Compute/virtualMachines/delete. It also has another assigned role that grants VM deletion on the same scope. Can it delete a VM, ignoring locks and deny assignments?
- No, because VM deletion is always a data-plane action
- Only if the account is assigned Global Administrator in Microsoft Entra
- Yes, because NotActions is not an explicit deny against permissions from another role
- No, because NotActions overrides all other roles
Explanation
Correct Answer
C:Azure RBAC combines allowed operations from role assignments. NotActions subtracts from that one role definition, but a separate applicable role may still grant the deletion action.
Incorrect Answers
A:Deleting an Azure VM is a resource management-plane action, not blob or VM guest data access. The premise about data actions is false.
B:Entra tenant directory administration is not a prerequisite for deleting a VM with proper Azure RBAC rights. Resource authorization is evaluated at Azure scope.
D:NotActions has no universal-deny behavior; it only affects the set granted by its own role. Another assignment can restore the same action.
Question 70
A custom role grants Microsoft.Compute/virtualMachines/read and Microsoft.Compute/virtualMachines/start/action but has no delete action. A VM operator can start the VM and see properties but requests to delete it fail. Which permission change would authorize only the extra deletion operation if the business now approves it?
- Add Microsoft.Storage/storageAccounts/blobServices/containers/blobs/read to DataActions
- Add the VM deletion action to NotActions rather than Actions
- Add the VM delete operation under DataActions instead of Actions
- Add the specific Microsoft.Compute/virtualMachines/delete action to the custom role and review its scope
Explanation
Correct Answer
D:Azure custom roles enumerate management actions, so extending the Actions allowlist with the VM delete operation grants the targeted capability. The assignment must remain appropriately scoped.
Incorrect Answers
A:Blob data read is unrelated to deleting an Azure VM resource and would not satisfy the denied Compute management action.
B:NotActions removes an operation from the custom role's own allowed management actions; it cannot grant the newly approved delete capability.
C:Deleting the VM resource is an ARM management-plane operation. The custom role must add the matching management Action.
Microsoft Azure Administrator (AZ-104) Practice Tests