Question 70 A resource group contains a locked critical database and expendable test VMs. Where should a delete lock go to protect only the database? A generic CostCenter tag The entire resource group The whole subscription The database resource, if the service supports the intended scope Explanation Correct Answer D:A resource-level lock avoids blocking deletion of unrelated test VMs. Incorrect Answers A:Tags cannot enforce deletion protection. B:Group lock affects both database and disposable VMs. C:Subscription lock affects many unrelated resources.
Question 71 A network team wants to inspect why one backend never receives load-balanced connections. What should be reviewed? Inspect only the load balancer frontend public or private IP configuration Increase the load-balancing rule idle timeout without checking probe failures Change the health probe path without checking the application listener or NSGs Backend pool membership, probe health and NSG/host listener configuration Explanation Correct Answer D:Distribution depends on inclusion, health and reachability. Incorrect Answers A:A correct frontend address does not establish that the particular backend pool member is healthy and reachable. B:Idle timeouts govern existing connections and cannot repair an unhealthy or excluded backend. C:Probe-path changes alone do not establish pool membership, listener readiness and security-rule reachability.
Question 72 A Site Recovery setup has replication errors after a subnet was removed in the target region. What should be checked? Increase Recovery Services vault backup retention while leaving replication mappings broken Change the replication frequency without recreating the target subnet mapping Recovery target network mappings and the replication health details Reconfigure the source subnet NSG only and assume failover networking is ready Explanation Correct Answer C:Target network dependencies are needed for failover VMs to attach correctly. Incorrect Answers A:Backup retention cannot repair the target network resources on which Site Recovery depends. B:Replication timing does not restore a removed destination subnet or its recovery configuration. D:Changes in the primary region do not restore the missing target-region subnet assignment.
Question 73 A storage firewall allows selected subnets, but a new subnet is blocked even though both share a VNet. What should be reviewed? The precise subnet's network rule and applicable endpoint configuration The storage account's friendly display name Blob soft delete retention The VM's availability-set membership Explanation Correct Answer A:Storage VNet rules are scoped to permitted subnets, not implicitly the whole VNet. Incorrect Answers B:Names do not authorize additional subnets. C:Retention does not control allowed source networks. D:Availability sets do not authorize storage network access.
Question 74 A VNet must have an address space large enough for 200 hosts plus future growth. What planning concern is primary? Use one /30 subnet for all hosts Use DNS TTL as a substitute for IP capacity Allocate overlapping subnets within one VNet Choose a nonoverlapping CIDR range and subnet sizes that leave capacity Explanation Correct Answer D:Address planning must fit hosts and avoid future peering collisions. Incorrect Answers A:/30 cannot address hundreds of hosts. B:TTL has no bearing on address space size. C:Overlapping subnet spaces create conflicts.
Question 75 A VM service needs outbound access to Azure platform endpoints without manually tracking IP ranges. What NSG option supports broad service classification? Use a blob object's index tags Use supported Azure service tags in rules instead of hardcoded ranges Use a random ASG name without members Use subscription Owner instead of network rules Explanation Correct Answer B:Service tags track Microsoft-managed service IP address prefixes. Incorrect Answers A:Blob tags cannot match network destinations. C:An empty ASG does not represent service IPs. D:Owner cannot replace packet filtering.
Question 76 A custom domain is added to an App Service app. What DNS step is normally required to map traffic? Enable blob versioning Create a VNet route table only Delete the app's default hostname Create the appropriate DNS mapping and complete domain ownership validation Explanation Correct Answer D:Custom hostname binding requires DNS configuration and validation. Incorrect Answers A:Blob versions do not map app domains. B:Routes do not establish public DNS hostname binding. C:The default hostname is not removed to validate custom domain.
Question 77 An internal application should accept traffic only from private clients, spread across several backend VMs. Which Azure load balancer fits? Public Load Balancer with an internet frontend Internal Azure Load Balancer with a private frontend IP Azure DNS TXT record only Azure Storage object replication Explanation Correct Answer B:An internal load balancer distributes traffic inside private network boundaries. Incorrect Answers A:A public frontend violates the internal-only goal. C:DNS records do not health-probe or distribute TCP connections. D:Storage replication is not a load balancer.
Question 78 An App Service app cannot reach on-premises resources through its configured VNet. What infrastructure dependency should be examined? Only the app's public custom domain VNet integration subnet, routing and hybrid connectivity from that VNet Only Azure Files snapshots Only the user's directory license state Explanation Correct Answer B:Outbound path requires correctly configured network routes to the target. Incorrect Answers A:Public hostname does not provide outbound hybrid routing. C:Snapshots are unrelated to connectivity. D:User licenses do not build a network route.
Question 79 A team needs an email and webhook when a VM CPU metric remains above 90% for ten minutes. What should be configured? An Azure Policy deny rule for large VMs An Entra group licensing rule A blob lifecycle tiering rule Azure Monitor metric alert rule with action group Explanation Correct Answer D:Metric alerts evaluate thresholds and action groups deliver notifications. Incorrect Answers A:Policy deployment enforcement does not notify on observed CPU spikes. B:Licensing does not monitor CPU. C:Tiering cannot trigger VM metric alerts.
Question 80 A VM is moving to another region and an engineer assumes a simple resource-group move changes its region. What is the correction? Changing the resource group name relocates the VM Adding a second NIC automatically moves the VM Rotating the VM login password changes its region Use a supported regional migration or replication/redeployment approach Explanation Correct Answer D:Ordinary resource-group moves do not change a resource's region. Incorrect Answers A:Resource group names are not VM region migration operations. B:An extra NIC does not relocate compute. C:Credentials do not control resource placement.