Microsoft AZ-104 Practice Test 7
Question 60
A deployment uses the mutable latest tag from a private Azure Container Registry. An engineer requires the next deployment to use precisely the previously approved image even after another latest tag push. Which image reference should be used?
- The repository login server without any image identifier
- The latest tag, assuming tags are immutable by default
- The approved image manifest digest, referenced as repository@sha256:...
- The registry's Azure resource group name as an image identifier
Explanation
Correct Answer
C:An OCI manifest digest is content-addressed and uniquely identifies the published image artifact. Deploying with repository@sha256:digest selects the approved version even when a mutable tag like latest is later repointed.
Incorrect Answers
A:A registry login server identifies the registry endpoint but not one immutable application artifact. Without an image version, the runtime cannot guarantee it will pull the build that was approved.
B:Tags such as latest can be reused for newer image manifests. If a tag moves, a deployment referring only to that tag may retrieve different bytes than the version previously tested.
D:The Azure resource group organizes the registry resource, not the image manifest inside it. That name cannot select a precise container version for deployment.
Question 61
A Container Apps service has two active revisions. Operations wants to send a small fraction of incoming traffic to the new revision before increasing rollout. Which feature should be used?
- Weighted traffic splitting between revisions
- A storage lifecycle rule
- A VM availability set
- An Entra security group's membership rule
Explanation
Correct Answer
A:Container Apps supports multiple active revisions and ingress traffic weights in suitable revision modes. Route a small share to the new version, observe health and error rates, then adjust weights as confidence grows.
Incorrect Answers
B:Blob lifecycle policies transition or delete storage objects based on configured conditions, not shift HTTP requests between application revisions.
C:Availability sets help distribute VM hardware fault and update domains. They do not govern Container Apps revision routing.
D:Directory group membership chooses principals, not percentage-based HTTP ingress distribution across running container revisions.
Question 62
An Azure App Service web app needs more instances at peak load without changing the compute size of each instance. Which operation should the operator perform?
- Scale out the App Service plan
- Scale up the plan to a larger worker SKU
- Change the web app's DNS label only
- Add a storage blob snapshot
Explanation
Correct Answer
A:Scaling out changes the number of worker instances available to supported apps on the plan. Configure autoscale conditions and limits while recognizing that plan pricing and instance quota are affected.
Incorrect Answers
B:Scaling up increases the resources of individual workers or changes the plan tier. It is a vertical change and does not directly meet the instance-count requirement.
C:A custom DNS label changes how clients find the app but does not add compute workers or response capacity.
D:Blob snapshots preserve storage data; they cannot add App Service serving instances when traffic increases.
Question 63
A development team wants a production App Service with a staging deployment slot so the new build can be warmed up and swapped with minimal service interruption. What prerequisite should be verified?
- The App Service plan supports deployment slots at its selected tier
- The app must have a VM availability set
- All site data must be placed in archive blob tier
- The tenant must disable Microsoft Entra groups
Explanation
Correct Answer
A:Deployment slots are available only on eligible App Service plan tiers. Verify support and slot count, deploy to staging, validate slot-specific settings, and plan how the swap handles warmed instances and configuration.
Incorrect Answers
B:Availability sets are VM placement resources and do not enable App Service slots. App Service has its own managed plan features.
C:Archive blob tier is offline storage and is not a prerequisite for a web app slot swap. It would slow any required content retrieval.
D:Entra groups manage identities. Disabling them would break organizational authorization and is unrelated to slot availability.
Question 64
A company maps www.contoso.example to its Azure App Service and needs browsers to use HTTPS with an organization-approved certificate. What must the admin configure?
- Custom domain DNS validation/mapping plus a valid TLS certificate binding
- Only an NSG on the storage account
- A management group tag called HTTPS
- Azure Backup snapshots on every VM
Explanation
Correct Answer
A:App Service needs proof of the custom hostname through the required DNS records and the hostname mapping, followed by a supported TLS certificate and binding. Validate certificate renewal and HTTPS-only configuration.
Incorrect Answers
B:NSGs on unrelated storage networks cannot prove ownership of the web hostname or provide an HTTPS server certificate to App Service.
C:A management group tag is metadata and does not configure DNS records, certificate issuance or TLS termination for a web app.
D:VM backup snapshots protect disk data and are unrelated to an App Service custom hostname and transport security configuration.
Question 65
An App Service must call a private API reachable through a virtual network. The requirement concerns outbound connectivity from the web app, not private inbound user access. What feature should be evaluated?
- App Service virtual network integration
- A public anonymous storage container
- An Azure DNS alias record alone
- A VM serial console session
Explanation
Correct Answer
A:Regional VNet integration provides outbound access from a supported App Service into networks with appropriate routing and DNS. Evaluate subnet delegation, plan tier and target endpoint access; it is not itself a private inbound endpoint.
Incorrect Answers
B:Anonymous storage changes data exposure and does not route outbound web app traffic to private API addresses.
C:DNS name resolution may be necessary but is not sufficient: resolving a private name does not establish a reachable network path for App Service.
D:A VM serial console is a recovery channel for supported VMs and does not configure App Service outbound traffic.
Question 66
An App Service administrator needs to back up web app content and supported configuration for recovery after an application change. Which feature should be planned?
- App Service backup where supported by the plan and storage requirements
- A network security group flow log
- An Azure Policy initiative only
- An Azure VM host encryption flag
Explanation
Correct Answer
A:App Service backup can create recoverable application backups under supported tiers and configurations. Configure destination storage, schedule or on-demand operation, and test restore against the app's real dependencies.
Incorrect Answers
B:NSG flow logs capture networking events, not a restorable web application content/configuration backup. They cannot recover an accidentally overwritten site deployment.
C:Policy initiatives evaluate governance and do not contain a runnable web-app snapshot. Compliance status is not a backup copy.
D:VM host encryption protects eligible VM data paths, not App Service configuration recovery. The application is a managed web service rather than the referenced VM.
Question 67
A new Azure VNet is planned as 10.20.0.0/16, but its required subnet is entered as 10.21.0.0/24. What should the administrator correct?
- Expand the subnet by changing it to /8 without evaluating overlaps
- Choose a subnet CIDR inside the VNet address space, such as 10.20.1.0/24
- Enable public anonymous access on the VNet
- Use a subscription-wide role assignment to change IP arithmetic
Explanation
Correct Answer
B:A subnet prefix must fall entirely within a configured VNet address space and must not overlap another subnet. 10.20.1.0/24 is within 10.20.0.0/16, unlike the proposed 10.21.0.0/24.
Incorrect Answers
A:Changing the subnet to a much larger prefix without checking address overlap is not a valid correction. It could encompass networks outside the VNet and violate subnet allocation rules.
C:Azure VNets do not use anonymous public-access flags to bypass their address space rules. Network topology validation must be correct before resource deployment.
D:RBAC changes which users may edit the VNet but cannot make an out-of-range subnet address belong to a different CIDR block.
Question 68
Two Azure VNets in different resource groups must communicate privately without a VPN gateway. Their IP ranges do not overlap. Which feature should the engineer configure?
- VNet peering between the networks
- An Azure Files snapshot
- A Microsoft Entra security group
- A Cost Management budget
Explanation
Correct Answer
A:VNet peering connects nonoverlapping virtual networks over Microsoft's private backbone with appropriate routing and security rules. Configure peering direction and verify NSGs, DNS and traffic requirements.
Incorrect Answers
B:Snapshots capture point-in-time Azure Files data and cannot establish network routes between two VNets.
C:Entra security groups manage directory principals rather than address-space routing. They cannot replace peering connectivity.
D:A budget measures spending and sends alerts; it has no function to route private packets between virtual network address spaces.
Question 69
VNet-A peers with VNet-B, and VNet-B peers with VNet-C. A developer assumes A can reach C through B by default. What is the correct expectation?
- Peering is automatically transitive across all three VNets
- A-to-C connectivity requires its own supported routing architecture; peering alone is nontransitive
- B automatically becomes an internet gateway
- A management lock on B enables forwarded traffic
Explanation
Correct Answer
B:VNet peering connections are nontransitive. For A-to-C communication, configure direct peering or a deliberately designed hub routing solution using supported forwarding and gateway features.
Incorrect Answers
A:Direct Azure VNet peering does not automatically create a transitive network between every peer of a peer. Assuming it does can leave a connectivity gap.
C:A peered VNet does not become an internet gateway for unrelated address spaces without intentional forwarding and routing configuration.
D:Management locks govern resource changes, not forwarding behavior. They cannot enable transitive network connectivity.
Question 70
A spoke VNet must use a central hub's virtual network gateway for on-premises connectivity. Which peering settings should be examined?
- Hub gateway transit and spoke use-remote-gateways settings, within their supported constraints
- A blob SAS expiry time
- Entra self-service password reset
- Azure Cost Management tags
Explanation
Correct Answer
A:A hub-spoke gateway transit design requires the hub to allow gateway transit and the spoke to use the remote gateway where supported. Check peer relationship, gateway prerequisites and route propagation; remote gateway use has constraints.
Incorrect Answers
B:Blob SAS timing controls delegated storage access. It has no effect on VNet gateway path selection for on-premises routing.
C:Password reset settings are part of identity recovery and do not expose the hub gateway to a spoke's routing table.
D:Cost tags classify resources but cannot enable peering gateway transit or inherited routing.