Question 110 A company places a latency-sensitive service near customers while also considering data residency. What should its region selection account for? Only the alphabetic order of region names Only the number of developers assigned to the project Choose only the Azure region with lowest storage pricing Network proximity and applicable data location obligations together Explanation Correct Answer D:A suitable Azure region must satisfy both user experience and governance constraints. Incorrect Answers A:Names do not indicate performance or legal suitability. B:Staffing does not determine network latency or permitted data locations. C:A lowest-cost region may create excessive user latency or fail the organization’s residency obligations.
Question 111 A financial application cannot fail merely because one datacenter building loses power within a supported Azure region. Which design has the relevant failure-domain separation? Run additional processes on one VM in one building Place the only data copy on a larger attached disk Assign multiple resource tags to the existing VM Deploy compatible resources redundantly across availability zones Explanation Correct Answer D:Separate zones can reduce the effect of a single-zone datacenter failure. Incorrect Answers A:Multiple processes still depend on the same host and site. B:Disk size does not create zone redundancy. C:Tags provide metadata, not physically independent capacity.
Question 112 A business must maintain a service even if the entire primary Azure region becomes unavailable. What capability does its design need? A resource lock on the primary service Only redundant instances within one region A tested multi-region recovery or failover arrangement Only autoscale rules on a regional VM scale set Explanation Correct Answer C:Regional continuity requires an independent alternate deployment and a recovery mechanism. Incorrect Answers A:Deletion protection does not create a second-region service. B:Same-region replicas share the region outage boundary. D:Autoscaling cannot relocate a failed region automatically.
Question 113 A load-balanced application contains one unhealthy VM that should stop receiving new traffic. Which capability supports this outcome? A longer application hostname Health probes and traffic routing to healthy instances A backup policy without any traffic failover An accounting tag describing the VM owner Explanation Correct Answer B:Health-aware distribution can exclude a failed backend while healthy peers continue. Incorrect Answers A:Names do not signal runtime health to a load balancer. C:Backups can help recovery but do not divert live requests automatically. D:Metadata does not check application responsiveness.
Question 114 A service must keep serving customers while one instance receives maintenance. Which architecture assists? No deployment automation and no spare capacity Multiple replaceable instances with controlled rotation A single storage snapshot as the serving tier One single instance whose updates always restart the service Explanation Correct Answer B:Traffic can shift to healthy peers while one worker is unavailable for maintenance. Incorrect Answers A:Without alternatives, maintenance causes interruption. C:A snapshot is not a redundant live application worker. D:The only instance creates a maintenance outage.
Question 115 Two web servers provide redundancy, but users lose sessions when requests change servers. Which design improvement supports available service? Replace the load balancer with a static host file on every laptop Externalize shared session state to durable storage Keep all sessions only in the first server's local RAM Disable the second server during peak hours Explanation Correct Answer B:Sessions then survive movement between application instances. Incorrect Answers A:Manual fixed routing creates brittleness rather than shared state. C:Failure or traffic movement removes access to that session. D:Removing a replica reduces high availability.
Question 116 A team duplicates an application database to support high availability. Which issue must its design still address? Replication consistency and appropriate failover behavior The assumption that every replica is always fully current under any failure Using identical table display colors Sharing an administrator password among replicas as the sole mechanism Explanation Correct Answer A:Multiple copies need a well-defined state and recovery model, not just extra disks. Incorrect Answers B:Replication timing and modes can vary. C:Presentation choices do not ensure data consistency. D:Credentials do not define recovery or transaction ordering.
Question 117 An architect assumes every Azure resource SKU can be deployed across multiple availability zones in every region. What should be checked? A rule that any paid service is automatically zone-redundant Only whether the resource is in a resource group Regional and SKU support for the intended zone-redundant configuration Only whether a region has a short name Explanation Correct Answer C:Availability zone support varies by service, region and tier. Incorrect Answers A:Paid products may still require explicit compatible configurations. B:Logical grouping does not guarantee zoning. D:Naming convention does not establish zone support.
Question 118 A company spans a service across zones and asks whether it must still plan application failover. What is correct? Zone separation helps, but applications must handle failures and dependencies appropriately Application failover is unnecessary if a resource has a name Zones guarantee no application bug can cause downtime A zone deployment eliminates network latency between all users Explanation Correct Answer A:Infrastructure separation alone does not fix state, routing or application behavior. Incorrect Answers B:Resource identifiers do not implement continuity. C:Availability zones do not validate code. D:Zones address infrastructure fault domains rather than global latency.
Question 119 A transaction requires both an API and an identity service to be healthy. Why can its end-to-end availability differ from either component's SLA? An SLA replaces all architectural design work SLAs always add together to create perfect application uptime The combined service depends on all necessary components succeeding Authentication availability cannot affect a protected API Explanation Correct Answer C:Even high individual availability targets do not guarantee their composite uptime. Incorrect Answers A:Tenant redundancy and dependency design still matter. B:Combined dependencies can reduce effective end-to-end availability. D:If sign-in is required, identity outages can block access.
Question 120 A team wants to know when autoscale should add web instances. What evidence should guide the threshold? A contract that requests never fluctuate Only the VM resource display names Measured traffic, response times, utilization and capacity tests A universal CPU threshold guaranteed optimal for every application Explanation Correct Answer C:Workload behavior provides better scaling triggers than arbitrary guesses. Incorrect Answers A:Demand should be measured, not assumed constant. B:Names provide no capacity measurements. D:Appropriate signals vary by workload.
