Azure’s shift away from paired-region-only disaster recovery toward availability zones and zone-redundant services is reshaping how architects design for resilience. This is the fourth set (Questions 31–40) in our AZ-305 practice series. These AZ-305 business continuity questions cover Azure Site Recovery, backup strategies, high availability design, RPO/RTO calculations, and geo-redundant architectures.
Question 31: Designing VM Disaster Recovery Across Regions
A company runs a critical three-tier application on Azure Virtual Machines in the East US region. The business requires automated failover to a secondary region with an RPO under 15 minutes and an RTO under 1 hour. The VMs must fail over in a specific order: database first, then application tier, then web tier.
Which solution should you recommend?
A) Azure Backup with cross-region restore
B) Azure Site Recovery with a recovery plan that defines boot order groups
C) Custom scripts using Azure Automation to redeploy VMs from snapshots
D) Availability zones within the same region
👁 Reveal Answer
Correct Answer: B
Explanation: Azure Site Recovery (ASR) provides continuous block-level replication to a secondary region, achieving RPO under 15 minutes and RTO under 1 hour through automated failover. Recovery plans allow you to define boot order groups, ensuring the database VMs start first, followed by the application tier and web tier. Option A uses backup snapshots with RPO of hours (based on backup frequency), exceeding the 15-minute requirement. Option C requires custom development and maintenance with no SLA on failover time. Option D provides protection against datacenter failure within a region but does not protect against a full regional outage.
Question 32: Calculating Composite SLA
An architect designs a web application with the following Azure services in sequence: Azure Front Door (99.99%), Azure App Service (99.95%), and Azure SQL Database (99.99%). All three services must be operational for the application to work. What is the composite SLA, and what should the architect recommend to improve it?
A) Composite SLA is 99.93%. Deploy App Service across availability zones to raise its SLA.
B) Composite SLA is 99.99%. No changes needed.
C) Composite SLA is 99.95%. Add Azure SQL geo-replication.
D) Composite SLA is 99.93%. Replace Azure Front Door with Azure Traffic Manager.
👁 Reveal Answer
Correct Answer: A
Explanation: The composite SLA for serial dependencies is calculated by multiplying individual SLAs: 0.9999 × 0.9995 × 0.9999 = 0.9993 (99.93%). The weakest link is App Service at 99.95%. Deploying App Service with zone redundancy increases its SLA to 99.99%, which would raise the composite SLA to approximately 99.97%. Option B incorrectly states 99.99% — you cannot achieve a composite SLA higher than the lowest individual SLA in a serial chain. Option C identifies the wrong composite value and targets SQL (99.99%), which is not the weakest link. Option D changes the wrong component — Front Door already has 99.99% SLA.
Question 33: Designing Backup for Azure Files
A company uses Azure Files (SMB) for shared departmental documents. They need daily backups retained for 30 days, with the ability to restore individual files or entire shares. The backup solution must be native to Azure and require minimal configuration.
Which solution should you recommend?
A) Azure Backup with the Azure file share backup policy
B) Azure File Sync with cloud tiering enabled
C) Use AzCopy scheduled via Azure Automation to copy files to a secondary storage account nightly
D) Enable soft delete on the storage account and rely on snapshot management
👁 Reveal Answer
Correct Answer: A
Explanation: Azure Backup natively supports Azure file share backup with configurable daily backup schedules and retention policies up to years. It supports both full share restore and individual file-level restore from the Azure portal, meeting all requirements with minimal setup. Option B synchronizes files between on-premises and Azure but is not a backup solution — it does not provide point-in-time recovery. Option C creates copies but requires custom scripting, scheduling, and does not offer native file-level restore from a backup vault. Option D protects against accidental deletion but soft delete has a limited retention window (1–365 days) and does not provide structured backup policies or vault-based management.
Question 34: High Availability for Azure SQL with Zone Redundancy
A retail application uses Azure SQL Database on the Business Critical tier. The database must survive a datacenter failure within the region without data loss and with automatic failover in under 30 seconds. The solution must also provide a readable secondary for reporting queries.
Which configuration should you recommend?
A) General Purpose tier with zone redundancy enabled
B) Business Critical tier with zone redundancy enabled
C) Hyperscale tier with a named replica in another availability zone
D) Business Critical tier with auto-failover group to another region
👁 Reveal Answer
Correct Answer: B
Explanation: Business Critical tier with zone redundancy distributes replicas across availability zones within the region. It provides automatic failover with near-zero data loss (synchronous replication), failover within 30 seconds, and a built-in read-only endpoint for offloading reporting queries. Option A provides zone redundancy but the General Purpose tier does not include a readable secondary replica. Option C provides read scale-out but named replicas are in the same region without the same zone-redundant HA guarantees as Business Critical. Option D replicates across regions (useful for DR) but uses asynchronous replication with potential data loss — overkill for the in-region datacenter failure scenario described.
Question 35: Selecting RPO and RTO for Different Workloads
An enterprise classifies its workloads into three tiers. Tier 1 (payment processing) requires RPO of 0 and RTO under 1 minute. Tier 2 (internal CRM) accepts RPO of 1 hour and RTO of 4 hours. Tier 3 (archived reports) accepts RPO of 24 hours and RTO of 24 hours. The architect needs to match each tier to the most cost-effective Azure solution.
Which mapping is correct?
A) Tier 1: Availability zones with synchronous replication; Tier 2: Azure Site Recovery; Tier 3: Azure Backup with daily snapshots
B) Tier 1: Azure Site Recovery; Tier 2: Azure Backup with hourly snapshots; Tier 3: GRS storage
C) Tier 1: Azure Backup with continuous backup; Tier 2: Availability zones; Tier 3: Azure Site Recovery
D) Tier 1: GRS with RA-GRS; Tier 2: LRS with daily backups; Tier 3: Archive storage with no backup
👁 Reveal Answer
Correct Answer: A
Explanation: Tier 1 needs zero data loss and instant failover — availability zones with synchronous replication (e.g., zone-redundant SQL Business Critical) achieves RPO 0 and sub-minute failover. Tier 2 needs RPO of 1 hour and RTO of 4 hours — Azure Site Recovery provides continuous replication with RPO under 15 minutes and RTO under an hour, comfortably meeting both targets. Tier 3 accepts daily RPO/RTO — Azure Backup with daily snapshots is the cheapest option. Option B assigns ASR to Tier 1, but ASR uses asynchronous replication and cannot guarantee RPO of 0. Option C puts Azure Backup on Tier 1 — backup cannot achieve RPO 0 or sub-minute RTO. Option D assigns storage redundancy only, which lacks automated failover and restore orchestration.
Question 36: Designing Disaster Recovery for Azure Kubernetes Service
A company runs microservices on Azure Kubernetes Service (AKS) in West Europe. They need a disaster recovery strategy that can bring up the application in North Europe within 2 hours. The application is stateless, but it connects to Azure SQL Database and Azure Cosmos DB for persistent data. Cluster configuration is managed as code in a Git repository.
Which DR approach should you recommend?
A) Use Azure Site Recovery to replicate AKS nodes to North Europe
B) Maintain an active-passive AKS cluster in North Europe using GitOps for configuration, with Azure SQL auto-failover group and Cosmos DB multi-region reads
C) Take daily VM snapshots of AKS nodes and restore them in North Europe during failover
D) Use Azure Traffic Manager with both clusters active and running at full capacity at all times
👁 Reveal Answer
Correct Answer: B
Explanation: For stateless AKS workloads, an active-passive approach using GitOps (Flux or Argo CD) to deploy the same configuration from the Git repo to the DR cluster is the recommended pattern. The passive cluster can run minimal nodes and scale up during failover. Azure SQL auto-failover groups handle database DR, and Cosmos DB multi-region reads provide data availability in the secondary region. Option A is not supported — Azure Site Recovery does not replicate AKS clusters (AKS is a managed service, not raw VMs you control). Option C treats AKS nodes as pets rather than cattle, which contradicts Kubernetes design principles — nodes are ephemeral. Option D maintains full capacity in both regions, which is the most expensive option and unnecessary for a 2-hour RTO.
Question 37: Configuring Azure Backup Retention for Compliance
A legal firm requires that VM backups be retained daily for 30 days, weekly for 12 weeks, monthly for 12 months, and yearly for 7 years. The firm also needs the ability to restore VMs to any point within the last 5 days with minimal data loss.
Which Azure Backup configuration should you recommend?
A) Enhanced backup policy with hourly backups (every 4 hours) and a custom retention policy matching the daily/weekly/monthly/yearly requirements
B) Standard backup policy with daily backups only and 7-year retention
C) Azure Backup with snapshot-only recovery for 5 days and a standard retention policy for long-term
D) Azure Site Recovery for continuous replication plus Azure Backup for long-term retention
👁 Reveal Answer
Correct Answer: A
Explanation: Enhanced backup policy supports hourly backup frequency (every 4, 6, 8, or 12 hours), providing granular recovery points within the last 5 days for minimal data loss. It also supports tiered retention policies — daily for 30 days, weekly for 12 weeks, monthly for 12 months, and yearly for 7 years. Option B only takes daily backups, so the maximum data loss in the 5-day window could be up to 24 hours. Option C provides instant restore from snapshots but snapshot retention is limited and does not provide the granular hourly recovery points needed. Option D adds unnecessary complexity — ASR is for disaster recovery failover, not backup retention compliance.
Question 38: Designing High Availability with Availability Sets vs. Zones
An application team deploys a two-node SQL Server Always On availability group on Azure VMs. They need protection against both hardware rack failures and full datacenter outages within the region. The network latency between nodes must remain under 2 milliseconds for synchronous replication to work reliably.
Which deployment model should you recommend?
A) Deploy both VMs in the same availability set
B) Deploy VMs across two availability zones
C) Deploy both VMs in the same proximity placement group
D) Deploy VMs in different regions with asynchronous replication
👁 Reveal Answer
Correct Answer: B
Explanation: Availability zones protect against full datacenter failures within a region — each zone is a separate datacenter with independent power, cooling, and networking. Network latency between zones within a region is under 2 milliseconds round-trip, which supports synchronous SQL Server replication. Option A protects against rack-level hardware failures but does not protect against a full datacenter outage — all VMs in an availability set reside in the same datacenter. Option C minimizes latency by placing VMs close together but provides no fault-domain or availability guarantees. Option D introduces cross-region latency far exceeding 2 milliseconds, making synchronous replication impractical.
Question 39: Implementing Geo-Redundancy for Azure Cosmos DB
A travel booking platform uses Cosmos DB and serves customers globally. During a regional outage, the platform must continue accepting new bookings with no downtime. Bookings must not be lost even if an entire Azure region goes offline. The application can tolerate occasional write conflicts.
Which Cosmos DB configuration should you recommend?
A) Single-region write with automatic failover enabled
B) Multi-region writes (multi-master) with last-writer-wins conflict resolution
C) Single-region write with manual failover
D) Single-region deployment with periodic backup and restore
👁 Reveal Answer
Correct Answer: B
Explanation: Multi-region writes (multi-master) allow every region to accept write operations. If one region goes offline, other regions continue processing bookings with zero downtime. The last-writer-wins conflict resolution policy handles the occasional conflicts the application can tolerate. No bookings are lost because writes succeed in any available region. Option A supports automatic failover but there is a brief period of write unavailability during the failover process — the requirement specifies “no downtime.” Option C requires manual intervention during an outage, causing prolonged downtime. Option D risks significant data loss because periodic backups could be hours old.
Question 40: Designing Cross-Region Backup for Azure Key Vault
A company stores critical encryption keys and secrets in Azure Key Vault in the East US region. The security team wants to ensure that keys remain recoverable even if the entire East US region becomes permanently unavailable. They also require that no keys can be permanently deleted for at least 90 days.
Which approach should you recommend?
A) Enable soft-delete with a 90-day retention period and purge protection on the Key Vault
B) Enable soft-delete with purge protection, and rely on Microsoft’s built-in geo-replication of Key Vault to the paired region
C) Manually export keys and import them into a second Key Vault in another region on a weekly schedule
D) Use Azure Backup to back up the Key Vault to a Recovery Services vault in another region
👁 Reveal Answer
Correct Answer: B
Explanation: Azure Key Vault automatically replicates contents to a paired region within the same geography. During a regional failure, Microsoft fails over the Key Vault to the secondary region (read-only initially, then read-write if the outage is prolonged). Soft-delete with a 90-day retention period and purge protection ensures keys cannot be permanently deleted for 90 days. Together, these features provide both geo-redundancy and deletion protection. Option A enables deletion protection but does not address regional geo-redundancy — it only protects within the same region. Option C introduces operational overhead and risks stale key copies if the export schedule is missed. Option D is not a supported scenario — Azure Backup does not support Key Vault as a backup source.
Study Tips for AZ-305 Business Continuity
- Know the RPO/RTO of every Azure BC/DR service. ASR: RPO ~15 min, RTO ~1 hour. Azure Backup: RPO = backup frequency. Availability zones: RPO 0, near-instant failover. Match workload tiers to the right service.
- Composite SLA calculation is testable. Multiply SLAs for serial dependencies. Use redundancy (availability zones, multi-region) to raise individual component SLAs.
- Azure Site Recovery does NOT support PaaS services. ASR replicates VMs and physical servers. For SQL Database, use auto-failover groups. For AKS, use GitOps-based redeployment.
- Understand availability sets vs. availability zones. Sets protect against rack failures in one datacenter. Zones protect against full datacenter failures. Zone-to-zone latency is under 2ms.
- Key Vault geo-replication is automatic. Microsoft handles it. You just enable soft-delete and purge protection. Don’t waste time designing manual key export workflows.
Catch up on earlier sets: AZ-305 Questions 1–10: Cost Optimization and AZ-305 Questions 21–30: Data Storage. For official Microsoft guidance, review the AZ-305 Business Continuity learning path. Coming up next: Domain 4 — Design Infrastructure Solutions.