AZ-305 practice questions and answers
All 45 questions from Full Practice Test 1 for Designing Microsoft Azure Infrastructure Solutions, with the correct answer and a full explanation for each — including why the other options are wrong. Free to read, no signup.
What this set covers
Questions are weighted to match the official AZ-305 exam guide. The real exam is 40-60 (not published; Microsoft states "Most Microsoft Certification exams typically contain between 40-60 questions; however, the number can vary depending on the exam.") questions in 100 minutes with a pass mark of 700 / 1000.
- Design identity, governance, and monitoring solutions13 q · 28%
- Design data storage solutions10 q · 22%
- Design business continuity solutions8 q · 18%
- Design infrastructure solutions14 q · 32%
An on-premises application requires an SMB file share in Azure that Windows servers can mount directly. Which service should the design specify?
- AAzure Files with an SMB share✓
- BAzure Blob Storage with a container
- CAzure Data Lake Storage Gen2 only
- DAzure Queue Storage
Correct answer: A — Azure Files with an SMB share
Azure Files exposes SMB and NFS shares that operating systems can mount natively. Blob and Data Lake storage are object stores accessed over REST or their own drivers, and Queue Storage is a messaging service.
Microsoft Learn — Azure FilesAn application must reduce read load on Azure SQL Database for a heavily read session cache. Which component should the design add?
- AAzure Cache for Redis✓
- BAn additional database replica in the same elastic pool
- CAzure Data Factory
- DAzure Storage queues
Correct answer: A — Azure Cache for Redis
An in-memory cache removes repeated reads from the database entirely and is the standard fit for session state. Adding databases to an elastic pool does not create a cache, Data Factory orchestrates data movement, and queues buffer messages.
Microsoft Learn — Azure Cache for RedisAn architecture requires a hub-and-spoke network where all spoke traffic to the internet is inspected centrally. Which design should be used?
- AAzure Firewall in the hub with user-defined routes forcing spoke egress through it✓
- BA network security group on each spoke subnet only
- CA public IP on each spoke virtual machine
- DVirtual network peering with default routes only
Correct answer: A — Azure Firewall in the hub with user-defined routes forcing spoke egress through it
Central inspection requires both a firewall in the hub and user-defined routes that force spoke traffic through it, otherwise traffic takes the default internet path. Network security groups filter without inspecting, public IPs bypass the hub entirely, and peering alone does not redirect egress.
Microsoft Learn — Hub-spoke network topologyA design must allow an application to query relational data with T-SQL while the platform handles patching, backups, and high availability. Which option fits with the least management overhead?
- AAzure SQL Database✓
- BSQL Server on an Azure virtual machine
- CAzure Database for MySQL
- DAzure Cosmos DB for MongoDB
Correct answer: A — Azure SQL Database
Azure SQL Database is the fully managed PaaS relational service supporting T-SQL with platform-managed patching, backup, and availability. A SQL Server virtual machine leaves the operating system and engine to you, and the MySQL and MongoDB options are different query languages and engines.
Microsoft Learn — Azure SQL DatabaseAn application must access Azure Storage from a virtual network without any traffic leaving the Microsoft backbone, and the storage account must not be reachable publicly. Which design fits?
- AA private endpoint for the storage account with public network access disabled✓
- BA service endpoint with the storage firewall left open to all networks
- CA NAT gateway on the subnet
- DA public IP on the storage account
Correct answer: A — A private endpoint for the storage account with public network access disabled
A private endpoint gives the storage account a private IP in the virtual network, and disabling public access ensures it cannot be reached any other way. An open firewall leaves it publicly reachable, a NAT gateway provides outbound internet access, and storage accounts do not take public IPs in that sense.
Microsoft Learn — Private endpoints for Azure StorageWhich two are valid design choices to meet a near-zero RPO for an Azure SQL Database in a multi-region architecture? (Select TWO.)
- AActive geo-replication to a secondary region✓
- BA failover group with automatic failover policy✓
- CWeekly full backups only
- DExporting BACPAC files nightly
- EScaling up the primary database tier
Correct answer: A, B — Active geo-replication to a secondary region · A failover group with automatic failover policy
Active geo-replication maintains a continuously replicated readable secondary and failover groups add automatic failover with a stable connection endpoint. Weekly backups and nightly exports give RPOs of days, and scaling up improves performance rather than recoverability.
Microsoft Learn — Failover groupsAn application must remain available if a single datacentre within an Azure region fails. Which design element addresses this?
- ADeploy across availability zones with a zone-redundant load balancer✓
- BDeploy two virtual machines in the same availability zone
- CUse a single virtual machine with premium disks
- DEnable Azure Backup daily
Correct answer: A — Deploy across availability zones with a zone-redundant load balancer
Availability zones are physically separate datacentres within a region, so spreading instances across them with a zone-redundant front end survives the loss of one. Two instances in one zone share the same failure domain, a single virtual machine has no redundancy, and backups are recovery rather than availability.
Microsoft Learn — Availability zonesAn application design must deliver static content globally with low latency and offload it from the origin. Which approach fits?
- AServe static assets from Azure Front Door or a CDN in front of blob storage✓
- BServe all assets from a single virtual machine in one region
- CEmbed all assets in the application binary
- DUse Azure Service Bus to deliver images
Correct answer: A — Serve static assets from Azure Front Door or a CDN in front of blob storage
Caching static content at edge locations in front of object storage cuts latency for global users and removes that load from the origin. A single-region virtual machine gives poor global latency, embedding assets bloats deployments, and Service Bus is a messaging service.
Microsoft Learn — Static content hosting patternA design must ensure virtual machines can be administered without exposing RDP or SSH ports to the internet. Which service should be included?
- AAzure Bastion✓
- BA jump box with a public IP and an open RDP port
- CA network security group allowing RDP from any source
- DAzure Front Door
Correct answer: A — Azure Bastion
Bastion provides browser-based RDP and SSH over TLS from the Azure portal with no public IP or open management port on the target virtual machines. A public jump box and an open RDP rule are the exposure being removed, and Front Door publishes HTTP applications.
Microsoft Learn — Azure BastionA design must give a virtual machine scale set the ability to add instances automatically when the queue depth of a Service Bus queue grows. Which mechanism should be specified?
- AAn autoscale rule based on a custom metric for the queue length✓
- BA manual instance count change during business hours
- CA resource lock preventing scale-in
- DA larger virtual machine size with no scaling rules
Correct answer: A — An autoscale rule based on a custom metric for the queue length
Autoscale can evaluate a metric other than CPU, so a rule driven by queue length grows the fleet when backlog builds and shrinks it when the queue drains. Manual changes do not react to load, locks prevent modification, and a bigger fixed instance still has no elasticity.
Microsoft Learn — Autoscale with custom metricsAn application stores millions of JSON documents and needs single-digit millisecond reads with global distribution and multi-region writes. Which service should the design use?
- AAzure Cosmos DB✓
- BAzure SQL Database
- CAzure Table storage in a single region
- DAzure Files
Correct answer: A — Azure Cosmos DB
Cosmos DB is the globally distributed NoSQL database with turnkey multi-region writes and single-digit millisecond latency backed by SLA. Azure SQL is relational and does not offer multi-region writes in the same way, single-region Table storage is not globally distributed, and Azure Files is a file share.
Microsoft Learn — Azure Cosmos DBWhich two components should a design include to centralise logs and metrics from virtual machines, App Service, and Azure SQL for querying with KQL? (Select TWO.)
- AA Log Analytics workspace✓
- BDiagnostic settings on each resource sending data to the workspace✓
- CAn Azure Storage account with static website hosting
- DAn Azure Service Bus namespace
- EAn Azure Batch pool
Correct answer: A, B — A Log Analytics workspace · Diagnostic settings on each resource sending data to the workspace
The workspace is the query target and diagnostic settings are the mechanism that routes platform logs and metrics into it. Static website hosting, Service Bus, and Batch serve unrelated purposes.
Microsoft Learn — Diagnostic settingsA design must alert the operations team when the average CPU of a virtual machine scale set exceeds 80% for 15 minutes. Which component should be specified?
- AAn Azure Monitor metric alert rule with an action group✓
- BAn Azure Policy audit effect
- CA resource lock
- DA Log Analytics workspace retention setting
Correct answer: A — An Azure Monitor metric alert rule with an action group
A metric alert evaluates the metric against a threshold over a window and an action group delivers the notification. Policy audits configuration, locks prevent changes, and retention settings only control how long data is kept.
Microsoft Learn — Metric alertsAn Azure virtual machine application must access Azure Key Vault without any stored credentials. What should the design specify?
- AA managed identity for the virtual machine with an appropriate Key Vault role assignment✓
- BA service principal with a client secret in the application configuration file
- CA shared access signature embedded in the code
- DThe subscription owner's credentials
Correct answer: A — A managed identity for the virtual machine with an appropriate Key Vault role assignment
A managed identity is issued and rotated by the platform, so the application authenticates with no secret to store or leak. A client secret in a config file is exactly the credential being removed, SAS tokens apply to storage, and using owner credentials is a severe privilege violation.
Microsoft Learn — Managed identitiesA design must protect Azure virtual machine backups from deletion by a compromised administrator account. Which feature should be included?
- AImmutable vaults and soft delete in the Recovery Services vault, with multi-user authorisation✓
- BStoring backups in the same virtual machine's data disk
- CA tag marking backups as protected
- DIncreasing backup frequency
Correct answer: A — Immutable vaults and soft delete in the Recovery Services vault, with multi-user authorisation
Immutability, soft delete, and multi-user authorisation together stop a single compromised identity from destroying recovery points. Storing backups on the protected machine defeats the purpose, tags carry no enforcement, and more frequent backups do not prevent deletion.
Microsoft Learn — Security features for Azure BackupA design requires that virtual machine images used across the organisation are versioned, replicated to multiple regions, and shared between subscriptions. Which service fits?
- AAzure Compute Gallery✓
- BAzure Container Registry
- CAzure Artifacts
- DAzure Blob Storage
Correct answer: A — Azure Compute Gallery
Azure Compute Gallery stores versioned virtual machine image definitions and replicates them across regions with sharing across subscriptions and tenants. Container Registry holds container images, Azure Artifacts holds software packages, and blob storage would require you to build all of this yourself.
Microsoft Learn — Azure Compute GalleryA cost design must attribute Azure spend to individual applications that each span several resource groups. Which approach should be specified?
- AA required tag for the application, enforced with Azure Policy, and cost analysis grouped by that tag✓
- BOne subscription per resource group
- CManual monthly review of each resource blade
- DA budget alert on the whole tenant only
Correct answer: A — A required tag for the application, enforced with Azure Policy, and cost analysis grouped by that tag
Tags applied consistently and enforced by policy let Cost Management group spend by application regardless of resource group boundaries. Subscription sprawl is a heavy structural change, manual review does not scale, and a tenant-wide budget gives no per-application breakdown.
Microsoft Learn — Cost analysis with tagsA company must require multi-factor authentication for all administrators when they sign in from outside the corporate network. Which feature should the design use?
- AMicrosoft Entra Conditional Access with a named location condition✓
- BAzure Policy assignment on the subscription
- CA network security group rule
- DAzure Firewall application rules
Correct answer: A — Microsoft Entra Conditional Access with a named location condition
Conditional Access evaluates signals such as location and role at sign-in and can require MFA only when the conditions match. Azure Policy governs resource configuration, and network security groups and firewalls filter traffic rather than controlling authentication.
Microsoft Learn — Conditional AccessA design requires a large-scale analytics store with hierarchical namespace support for Spark workloads. Which option fits?
- AAzure Data Lake Storage Gen2✓
- BAzure Table Storage
- CAzure Cache for Redis
- DAzure Files premium
Correct answer: A — Azure Data Lake Storage Gen2
Data Lake Storage Gen2 adds a hierarchical namespace to blob storage, which is what makes directory operations efficient for analytics engines. Table Storage is a key-value store, Redis is an in-memory cache, and Azure Files is a general purpose file share.
Microsoft Learn — Azure Data Lake StorageExternal partner users must access a specific Azure application without accounts being created in the corporate directory manually. Which design fits?
- AMicrosoft Entra External ID with B2B collaboration and guest invitations✓
- BCreating a shared member account for the partner organisation
- CGiving partners VPN access to the corporate network
- DPublishing the application anonymously
Correct answer: A — Microsoft Entra External ID with B2B collaboration and guest invitations
B2B collaboration lets partners sign in with their own identity while access is governed by the host tenant, with no manual account provisioning. A shared account removes attribution, VPN access grants far more than the application, and anonymous publishing removes access control entirely.
Microsoft Learn — B2B collaborationA design must prevent anyone from creating resources outside two approved Azure regions across every subscription in a management group. Which service should be used?
- AAzure Policy with an allowed locations policy assigned at the management group✓
- BAzure RBAC role assignments per resource group
- CResource locks on each resource
- DAzure Monitor alert rules
Correct answer: A — Azure Policy with an allowed locations policy assigned at the management group
Azure Policy evaluates resource properties at creation and can deny anything outside the allowed regions, and assigning it at the management group covers all child subscriptions. RBAC controls who may act rather than what properties are allowed, locks prevent deletion, and alerts only notify.
Microsoft Learn — Azure Policy overviewWhich two are valid reasons to organise subscriptions under management groups in a governance design? (Select TWO.)
- APolicy and RBAC assignments can be inherited by all child subscriptions✓
- BGovernance can be applied consistently as new subscriptions are added✓
- CManagement groups reduce the cost of virtual machines
- DManagement groups replace the need for resource groups
- EManagement groups provide network connectivity between subscriptions
Correct answer: A, B — Policy and RBAC assignments can be inherited by all child subscriptions · Governance can be applied consistently as new subscriptions are added
Management groups exist to make policy and role assignments inherit down the hierarchy so governance applies consistently, including to subscriptions added later. They do not affect pricing, they do not replace resource groups as a deployment boundary, and they provide no networking.
Microsoft Learn — Management groupsWhich two capabilities should a design include so that blob data is protected against accidental deletion and overwrite? (Select TWO.)
- ABlob versioning✓
- BSoft delete for blobs and containers✓
- CStatic website hosting
- DStorage account firewall rules only
- EChanging the redundancy option to LRS
Correct answer: A, B — Blob versioning · Soft delete for blobs and containers
Versioning keeps prior versions when a blob is overwritten and soft delete retains deleted blobs for a retention period, so both accidents are recoverable. Static website hosting is a feature for serving content, firewall rules control network access, and LRS reduces redundancy rather than protecting against deletion.
Microsoft Learn — Blob soft deleteWhich statement best describes the difference between Azure Backup and Azure Site Recovery in a continuity design?
- ABackup restores data from point-in-time copies, while Site Recovery replicates workloads for failover to another region✓
- BBackup replicates continuously and Site Recovery takes daily snapshots
- CBoth provide identical capabilities and only one is ever needed
- DSite Recovery is only for on-premises physical servers
Correct answer: A — Backup restores data from point-in-time copies, while Site Recovery replicates workloads for failover to another region
The two address different objectives: Backup protects against data loss and corruption with restorable copies, and Site Recovery keeps a warm replica for regional failover. Their roles are not reversed, they are complementary rather than redundant, and Site Recovery supports Azure virtual machines as well as on-premises workloads.
Microsoft Learn — Azure Backup overviewA design must expose a web application to the internet with TLS termination, path-based routing, and a web application firewall. Which service should be selected?
- AAzure Application Gateway with WAF✓
- BAzure Load Balancer standard
- CAzure Traffic Manager
- DAzure Private Link
Correct answer: A — Azure Application Gateway with WAF
Application Gateway is the layer 7 load balancer with TLS termination, URL path routing, and an integrated WAF. Azure Load Balancer works at layer 4, Traffic Manager routes at the DNS layer, and Private Link exposes services privately rather than publishing them publicly.
Microsoft Learn — Application GatewayA legacy application requires a specific Windows Server version and custom kernel-level agents. Which compute option should the design specify?
- AAzure Virtual Machines✓
- BAzure App Service
- CAzure Functions
- DAzure Container Apps
Correct answer: A — Azure Virtual Machines
Full control over the operating system version and low-level agents requires infrastructure as a service, which is what virtual machines provide. App Service, Functions, and Container Apps are managed platforms that do not expose that level of the stack.
Microsoft Learn — Choose an Azure compute serviceA storage design must survive the loss of an entire Azure region with a readable copy available. Which redundancy option should be selected?
- ARead-access geo-redundant storage✓
- BLocally redundant storage
- CZone-redundant storage
- DPremium file storage
Correct answer: A — Read-access geo-redundant storage
RA-GRS replicates to a secondary region and exposes a read endpoint there, which is what survives a regional loss with read access. LRS keeps copies in one datacentre, ZRS spans zones within one region, and premium file storage is a performance tier rather than a redundancy option.
Microsoft Learn — Storage redundancyA stateless web application must fail over between two regions with automatic health-based routing at the DNS layer. Which service should the design include?
- AAzure Traffic Manager with priority routing and endpoint monitoring✓
- BAn internal load balancer in each region
- CA virtual network peering between the regions
- DAzure Bastion in both regions
Correct answer: A — Azure Traffic Manager with priority routing and endpoint monitoring
Traffic Manager is a DNS-based global router that monitors endpoint health and directs clients to the highest priority healthy region. Internal load balancers distribute within a region, peering provides connectivity rather than routing decisions, and Bastion is a management access service.
Microsoft Learn — Traffic Manager routing methodsData in an Azure Storage account must be encrypted with a key the customer controls and can revoke. What should the design specify?
- ACustomer-managed keys in Azure Key Vault linked to the storage account✓
- BMicrosoft-managed keys with no configuration
- CClient-side base64 encoding
- DA shared access signature with a short expiry
Correct answer: A — Customer-managed keys in Azure Key Vault linked to the storage account
Customer-managed keys in Key Vault give control over rotation and revocation while the platform continues to encrypt at rest. Microsoft-managed keys offer no such control, base64 is an encoding rather than encryption, and SAS tokens grant access rather than encrypting data.
Microsoft Learn — Customer-managed keys for storageWhich two considerations should drive the choice between Azure Kubernetes Service and Azure App Service for a web API? (Select TWO.)
- AWhether the team needs fine-grained control over orchestration and networking✓
- BWhether the operational burden of running a cluster is acceptable✓
- CThe colour scheme of the Azure portal
- DThe number of resource groups in the subscription
- EThe tenant's display name
Correct answer: A, B — Whether the team needs fine-grained control over orchestration and networking · Whether the operational burden of running a cluster is acceptable
The real trade-off is control versus operational cost: AKS offers deep configurability but requires cluster operations, while App Service abstracts that away. Portal styling, resource group counts, and tenant names have no bearing on the decision.
Microsoft Learn — Choose a compute serviceAn architecture must decouple a front end from a back end so that a burst of requests does not overwhelm processing. Which component should be added?
- AAzure Service Bus queue with competing consumers✓
- BA larger front-end App Service plan
- CA second public IP address
- DAzure DNS private zones
Correct answer: A — Azure Service Bus queue with competing consumers
A queue absorbs the burst and lets consumers process at their own rate, which is the queue-based load levelling pattern. Scaling the front end does not protect the back end, extra addresses do not buffer work, and private DNS zones handle name resolution.
Microsoft Learn — Queue-based load levellingA business continuity design must be validated without affecting production. Which practice should be included?
- ARegular test failovers in an isolated network, measuring actual RTO and RPO✓
- BAn annual tabletop discussion only
- CTrusting the documented recovery times
- DDeleting production resources to observe recovery
Correct answer: A — Regular test failovers in an isolated network, measuring actual RTO and RPO
Site Recovery test failover creates the recovery environment in an isolated network so the plan is exercised and timed without customer impact. Discussion and documentation produce no evidence, and destroying production is reckless rather than a test.
Microsoft Learn — Test failover in Site RecoveryWhich two are appropriate when designing for Azure SQL Database point-in-time recovery requirements? (Select TWO.)
- AConfigure the backup retention period to cover the required recovery window✓
- BConfigure long-term retention policies for backups that must be kept for years✓
- CDisable automated backups to save cost
- DRely on application-level exports run manually
- EStore the database on ephemeral OS disks
Correct answer: A, B — Configure the backup retention period to cover the required recovery window · Configure long-term retention policies for backups that must be kept for years
The retention period defines how far back point-in-time restore can reach, and long-term retention covers multi-year regulatory requirements. Automated backups cannot be disabled for Azure SQL Database and would be reckless to avoid, manual exports are unreliable, and ephemeral disks are not how the PaaS service stores data.
Microsoft Learn — Automated backups in Azure SQL DatabaseWhich service should a design use to publish a globally distributed web application with edge caching, WAF, and origin failover?
- AAzure Front Door with WAF policy✓
- BAzure Application Gateway in a single region
- CAzure Load Balancer with a public IP
- DAzure ExpressRoute
Correct answer: A — Azure Front Door with WAF policy
Front Door is the global entry point with edge presence, caching, WAF, and automatic origin failover across regions. Application Gateway is regional, Azure Load Balancer works at layer 4 within a region, and ExpressRoute is a private connectivity circuit.
Microsoft Learn — Azure Front DoorA design requires that application teams can deploy only approved resource types into their subscriptions. Which combination best expresses this?
- AAzure Policy with an allowed resource types definition, plus RBAC granting Contributor scoped to the resource group✓
- BOwner role at the subscription with a written policy document
- CResource locks on unapproved resource types
- DA budget alert when unapproved resources are created
Correct answer: A — Azure Policy with an allowed resource types definition, plus RBAC granting Contributor scoped to the resource group
Policy constrains what may be created and RBAC constrains where the team may create it, which together express the requirement technically. Owner plus documentation relies on goodwill, locks cannot target resources that do not exist yet, and budget alerts respond to cost rather than resource type.
Microsoft Learn — Azure Policy built-in definitionsA design must connect an on-premises datacentre to Azure with private, high-bandwidth connectivity that does not traverse the public internet. Which option applies?
- AAzure ExpressRoute✓
- BSite-to-site VPN over the internet
- CPoint-to-site VPN for each server
- DAzure Bastion
Correct answer: A — Azure ExpressRoute
ExpressRoute provides a dedicated private circuit through a connectivity provider with predictable bandwidth and latency. Site-to-site and point-to-site VPNs ride the public internet, and Bastion is a managed jump host for RDP and SSH.
Microsoft Learn — ExpressRoute overviewA design requires that access to production resources is reviewed by resource owners every 90 days, with unused access removed. Which capability should be included?
- AMicrosoft Entra access reviews✓
- BAzure Monitor workbooks
- CAzure Advisor cost recommendations
- DAzure Resource Graph queries
Correct answer: A — Microsoft Entra access reviews
Access reviews schedule recurring attestation campaigns and can automatically remove access that reviewers do not approve. Workbooks visualise telemetry, Advisor makes cost recommendations, and Resource Graph queries inventory without driving a review workflow.
Microsoft Learn — Access reviewsWhich two designs are appropriate for an event-driven workload that processes messages only when they arrive and must not cost anything while idle? (Select TWO.)
- AAzure Functions on the Consumption plan triggered by a queue✓
- BAzure Container Apps with a scale rule that scales to zero✓
- CA virtual machine scale set with a fixed instance count
- DAn App Service plan reserved at the largest tier
- EA dedicated AKS cluster with fixed node pools
Correct answer: A, B — Azure Functions on the Consumption plan triggered by a queue · Azure Container Apps with a scale rule that scales to zero
Consumption Functions and Container Apps scale rules both allow the workload to drop to zero instances so idle time costs nothing. Fixed scale sets, reserved App Service plans, and fixed AKS node pools all bill continuously.
Microsoft Learn — Azure Functions hosting optionsA design must ensure that a critical resource group cannot be deleted accidentally, even by an owner. What should be included?
- AA CanNotDelete resource lock on the resource group✓
- BRemoving all Contributor role assignments
- CA tag named doNotDelete
- DAn Azure Monitor alert on delete operations
Correct answer: A — A CanNotDelete resource lock on the resource group
A CanNotDelete lock blocks deletion regardless of role, which is exactly the accidental deletion protection required. Removing Contributor assignments does not stop owners, tags carry no enforcement, and an alert fires after the fact.
Microsoft Learn — Lock resourcesArchive data must be retained for seven years, is accessed at most once a year, and can tolerate hours of retrieval latency. Which storage tier should be chosen?
- AArchive access tier with a lifecycle management policy✓
- BHot access tier
- CPremium block blob
- DCool access tier only
Correct answer: A — Archive access tier with a lifecycle management policy
The archive tier is the cheapest storage for rarely accessed data and its rehydration latency of hours is acceptable here, with lifecycle policies moving data there automatically. Hot and premium tiers are priced for frequent low-latency access, and cool is more expensive than archive for seven-year retention.
Microsoft Learn — Blob access tiersA design must host a containerised microservices application with automatic scaling, service discovery, and no cluster management. Which compute option fits best?
- AAzure Container Apps✓
- BAzure Kubernetes Service with self-managed node pools
- CVirtual machine scale sets running Docker manually
- DAzure Functions on the Consumption plan only
Correct answer: A — Azure Container Apps
Container Apps runs containerised microservices with built-in scaling, revisions, and service discovery without exposing a cluster to manage. AKS gives full Kubernetes control at the cost of managing it, virtual machine scale sets push all orchestration onto you, and Functions targets event-driven functions rather than long-running microservices.
Microsoft Learn — Azure Container AppsAdministrators should hold privileged roles only for the duration of an approved task, with justification and audit. Which solution should be designed?
- AMicrosoft Entra Privileged Identity Management with eligible role assignments✓
- BPermanent Owner assignments with a monthly review spreadsheet
- CA shared administrator account in a password vault
- DAzure Policy deny assignments
Correct answer: A — Microsoft Entra Privileged Identity Management with eligible role assignments
PIM makes roles eligible rather than active, requiring activation with justification, optional approval, and time limits, and it records the whole flow. Permanent assignments leave standing privilege, shared accounts destroy attribution, and deny assignments block actions rather than granting time-bound access.
Microsoft Learn — Privileged Identity ManagementA design must ensure that an Azure Storage account's data remains available if a single availability zone is lost, without cross-region replication. Which redundancy option fits?
- AZone-redundant storage✓
- BLocally redundant storage
- CGeo-redundant storage
- DRead-access geo-zone-redundant storage
Correct answer: A — Zone-redundant storage
ZRS synchronously replicates across three availability zones in one region, which meets zone resilience without leaving the region. LRS keeps all copies in one datacentre, and both geo-redundant options replicate to a second region, which the requirement excludes.
Microsoft Learn — Zone-redundant storageWhich design provides centralised security posture management and regulatory compliance scoring across Azure subscriptions?
- AMicrosoft Defender for Cloud✓
- BAzure Automation runbooks
- CAzure Site Recovery
- DAzure Cost Management budgets
Correct answer: A — Microsoft Defender for Cloud
Defender for Cloud provides secure score, recommendations, and regulatory compliance dashboards across subscriptions. Automation runs operational scripts, Site Recovery handles disaster recovery, and budgets track spend.
Microsoft Learn — Microsoft Defender for CloudA design requires an RPO of 15 minutes and an RTO of one hour for virtual machines running in a single region. Which service should be included?
- AAzure Site Recovery replicating to a secondary region✓
- BAzure Backup with daily backups only
- CAvailability zones within the same region
- DAzure Advisor recommendations
Correct answer: A — Azure Site Recovery replicating to a secondary region
Site Recovery replicates virtual machines continuously to another region and orchestrates failover, which is what meets a short RPO and RTO across regions. Daily backups give an RPO measured in a day, availability zones do not protect against regional loss, and Advisor makes recommendations without providing recovery.
Microsoft Learn — Azure Site RecoveryReady to try it under exam conditions?
Reading answers is not the same as recalling them with a clock running. Take the same 45 questions as a timed mock exam — 100 minutes, no feedback until you submit, then a score broken down by exam domain so you know what to study.
Start the timed AZ-305 test →