Question 350 A developer restarts a guest operating system inside an Azure VM. Which statement about the restart is accurate? It restarts every VM connected to the virtual network It restarts that VM's guest environment, not every resource in its resource group It automatically restarts all Azure regions in the subscription It restarts all VMs with the same tag Explanation Correct Answer B:Resource groups organize resources but do not synchronize guest restarts. Incorrect Answers A:A VNet does not impose shared guest lifecycle. C:A guest reboot has no subscription-wide region effect. D:Tags do not create a reboot broadcast.
Question 351 A critical application must survive a datacenter-zone failure within one supported Azure region. How should its compute be deployed? Place every replica in the same availability zone Use only different names for VMs in one zone Distribute instances across availability zones with appropriate traffic routing Use a single large VM with more CPU Explanation Correct Answer C:Zone placement reduces reliance on one datacenter zone. Incorrect Answers A:All replicas remain exposed to the same zone outage. B:Naming does not change the infrastructure failure domain. D:Capacity of one instance does not provide zone redundancy.
Question 352 A legacy two-VM service uses an availability set. What risk does that arrangement primarily reduce? A complete Azure region becoming unreachable Both VMs being affected by the same host or maintenance fault domain A software bug deployed identically to both VMs Use a VM Scale Set with one instance and assume that alone supplies fault-domain redundancy Explanation Correct Answer B:Availability sets distribute participating VMs across platform fault and update domains. Incorrect Answers A:An availability set does not span regions. C:Placement redundancy does not prevent identical application defects. D:A single instance remains a single failure point despite the fleet resource definition.
Question 353 A team believes an availability set guarantees VM placement in multiple separate availability zones. What distinction matters? A zone is just another label on a VM's resource group An availability set automatically creates three regional replicas Both features exclusively control outbound firewall rules Availability sets separate VM fault/update domains, whereas availability zones are physically separate datacenter zones Explanation Correct Answer D:The two mechanisms address different placement granularity. Incorrect Answers A:A zone is a deployment-location choice. B:Sets do not create instances or span regions. C:Neither is primarily a network filtering service.
Question 354 A web API on homogeneous Azure VMs needs more instances at peak load and fewer after midnight. What compute capability is designed for this? Virtual Machine Scale Sets with autoscaling rules One VM availability set containing no scale rules Azure Backup of one server A single VM with a static disk size Explanation Correct Answer A:A scale set manages multiple similar VM instances and supports capacity changes. Incorrect Answers B:Availability placement alone does not automatically change instance count. C:Backup aids recovery but does not scale active request processing. D:Disk capacity cannot adjust the number of running API instances.
Question 355 A company requires 20 identically configured worker VMs and coordinated updates. Which service is particularly suitable? Azure Policy alone Deploy individually configured VMs with divergent images and call them one uniform fleet Azure Files Azure Virtual Machine Scale Sets Explanation Correct Answer D:A scale set manages a repeatable model for a fleet of VM instances. Incorrect Answers A:Governance rules cannot replace provisioning and orchestration of 20 workers. B:Scale sets manage instance definitions and capacity as a coordinated group, not unmanaged unrelated machines. C:A shared file store does not create consistent compute instances.
Question 356 A batch team wants to scale a pool of identical workers based on queue pressure. What should determine the VMSS autoscale signal? A relevant monitored queue or workload metric with tuned thresholds The name of the VM image The number of tags on the worker resource group The age of the subscription Explanation Correct Answer A:Scaling should respond to pressure that actually predicts worker demand. Incorrect Answers B:Image metadata is not an operational pressure signal. C:Tag count does not measure queued tasks. D:Billing account tenure does not indicate queue load.
Question 357 A scale set hosting stateless HTTP endpoints loses an instance. How should application state be designed for recovery? Require user sessions to bind permanently to a single failed instance Store the only database on each instance's temporary disk Avoid load balancing so clients always target one worker Keep durable session and business state outside replaceable instances Explanation Correct Answer D:Instances can be recreated or scaled in without losing authoritative state. Incorrect Answers A:That prevents resilience to instance loss. B:Local-only state is lost with instance replacement. C:A single destination creates a service bottleneck and failure point.
Question 358 A scale set has eight web instances, but visitors still hit only one private IP. What additional component normally distributes inbound requests? A suitable Azure Load Balancer or application delivery service A storage account access tier A second resource group without networking settings Microsoft Purview sensitivity labels Explanation Correct Answer A:An ingress or load-balancing layer routes connections to healthy instances. Incorrect Answers B:Storage tier is unrelated to web connection distribution. C:Resource grouping does not spread traffic. D:Data classification cannot route HTTP requests.
Question 359 A shop adds three identical web VM instances instead of enlarging a single existing VM. What kind of scale action is this? Vertical scaling Data redundancy replication Subscription sharding Horizontal scaling Explanation Correct Answer D:Adding instances increases pool capacity rather than enlarging one machine. Incorrect Answers A:Vertical scaling changes CPU or memory of an individual instance. B:Storage copies are not extra application servers. C:Subscriptions are administrative and billing boundaries, not a synonym for more VMs.
Question 360 A specialized single-process service cannot be parallelized but needs more memory. Which change is vertical scaling? Move the resource to a different resource group Create another subscription Add identical web instances behind a load balancer Resize its VM to a supported higher-memory size Explanation Correct Answer D:Vertical scaling changes the resources of the individual instance. Incorrect Answers A:Administrative grouping does not resize compute. B:A subscription doesn't change guest RAM. C:Adding nodes is horizontal scaling and may not fit single-process constraints.
