These 14 questions test AZ-120 (Planning and Administering Microsoft Azure for SAP Workloads) the way the real exam does: a company scenario first, then several questions that hang off the same constraint set. Each scenario maps to one of the four official exam domains, weighted the same way Microsoft weights them. Answer them cold, then check the rationale — including why every wrong option is wrong — in the answer key at the end. If you want the domain breakdown and weightings behind these scenarios first, read the AZ-120 study guide.
How to use this:Read each scenario once, then answer every question under it before checking the answer key. Don't skip the rationale for questions you got right — the wrong-answer reasoning is where most of the exam-specific knowledge lives.
Scenario 1: Migrate SAP Workloads to Azure (Questions 1–4)
Exam weight: 25–30%
Scenario:Nordwind Manufacturing runs an on-premises SAP ECC landscape on Oracle and wants to move to Azure. The project has a fixed cutover date, and the CIO also wants the underlying database converted to SAP HANA before that date rather than as a separate later project. The infrastructure team is planning capacity across two Azure regions and wants the whole SAP estate — production and non-production — governed consistently. Part of the target architecture is an S/4HANA system that will run under RISE with SAP, which Nordwind's own BI reporting tools in Azure need to reach privately.
Given the fixed cutover date and the requirement to convert the database to HANA before it, which migration strategy should Nordwind Manufacturing choose?
A) Lift and shift
B) Lift-shift-migrate
C) Lift-shift-migrate to HANA
D) A greenfield S/4HANA reimplementation
Before deploying a large fleet of M-series VMs across two Azure regions for this landscape, which constraint should the infrastructure team check first?
A) Regional vCPU quota for the M-series family in the target subscription
B) The storage redundancy option for data disks
C) The commitment term for a future Reserved Instance purchase
D) The network security group rules for the app tier
Nordwind's BI reporting tools, running in its own Azure VNet, must reach the RISE-managed S/4HANA system without the traffic crossing the public internet. What should the team design?
A) A public endpoint on the RISE system with IP allow-listing
B) Private connectivity (such as VNet or ExpressRoute peering) to the RISE-managed environment
C) A site-to-site VPN over the public internet with NSG rules
D) Azure Front Door in front of the RISE system
Nordwind wants uniform Azure Policy enforcement and RBAC across every subscription in its SAP estate, with production and non-production kept separate. What should it implement?
A) A single flat subscription with resource groups per environment
B) Management groups aligned to environment tiers, with Azure Policy assigned at the management-group level
C) RBAC role assignments applied directly to each resource, with no policy layer
D) A single resource group holding all resources across every environment
Want more than 14 questions?
Try Free AZ-120 Questions in the Full Quiz Interface
Practice mode with instant explanations, covering every domain. No credit card required.
Start Free Practice →Scenario 2: Infrastructure for SAP Workloads (Questions 5–8)
Exam weight: 25–30%
Scenario:Fenwick Logistics is building a new SAP HANA production system on Azure. The SAP Basis team has already documented the application tier's throughput requirement, and the infrastructure team now needs to translate that into an actual Azure design — the right VM, the right storage for the HANA log volume, low latency between application and database tiers, and a repeatable way to deploy it all rather than a one-off manual build.
The application tier's throughput requirement is documented in SAPS. How should Fenwick Logistics choose the Azure VM?
A) Select any general-purpose VM with enough vCPUs to cover the workload
B) Select a VM from the SAP-certified VM types list that matches the required SAPS rating
C) Select the cheapest burstable B-series VM available
D) Select a VM based on GPU availability
The HANA log volume needs the lowest possible write latency on the chosen M-series VM. What should the team configure?
A) Standard HDD for the log volume
B) Write Accelerator enabled on a Premium SSD for the log volume
C) Blob Storage Archive tier for the log volume
D) Azure Files with the Cool access tier
The application-tier and database-tier VMs need minimal network latency between them within the same region. What should the team configure?
A) A proximity placement group for the application and database VMs
B) An Azure Load Balancer between the two tiers
C) ExpressRoute between the two VMs
D) A larger VM size only, with no other change
Fenwick wants a repeatable, Microsoft-supported automation framework built specifically for deploying SAP infrastructure, instead of a one-off manual build. What should it use?
A) The SAP Deployment Automation Framework (SDAF)
B) Manual deployment through the Azure Portal for each environment
C) Azure Migrate
D) A generic community Terraform module with no SAP-specific support
Scenario 3: High Availability and Disaster Recovery (Questions 9–11)
Exam weight: 20–25%
Scenario:Baltic Steel's production HANA database must keep running automatically if its primary node fails, and its cluster fencing must not depend on Azure Resource Manager API call limits during a regional API slowdown. Separately, the team must prove — not just assume — that its backup strategy actually meets its recovery point objective.
Baltic Steel needs automatic failover for the HANA database tier within a region if the primary node fails. What should it configure?
A) Deploy HANA on VMs in an availability set with no clustering software
B) Configure Pacemaker with HANA System Replication and fencing (STONITH)
C) Rely on a manual runbook to restart the failed VM
D) Configure only an Azure Load Balancer in front of the HANA VMs
The cluster's fencing mechanism must not depend on Azure Resource Manager API call rate limits. Which fencing mechanism should Baltic Steel choose?
A) The Azure fence agent, which uses ARM APIs
B) A STONITH Block Device (SBD) using iSCSI target VMs
C) No fencing, with manual intervention if split-brain occurs
D) Fencing via Azure Load Balancer health probes
Baltic Steel needs to prove its HANA backup strategy actually meets its recovery point objective. What should it do?
A) Configure Azure Backup for SAP HANA with a policy meeting the RPO, and periodically test restores
B) Rely solely on Azure Site Recovery replication with no separate backup policy
C) Take ad hoc manual disk snapshots with no defined policy
D) Skip restore testing, since Azure Backup validates backups automatically
Check your HA/DR reasoning before moving on
Take the Free AZ-120 Readiness Quiz
5 minutes, no signup, personalized to the domains you need most.
Check My Readiness →Scenario 4: Maintain SAP Workloads on Azure (Questions 12–14)
Exam weight: 20–25%
Scenario:Solna Retail Group's production SAP landscape is live and stable. The operations team now wants a single SAP-aware view of system health instead of piecing one together from generic dashboards, wants to trim compute cost on servers that run the same way every day, and wants to run planned maintenance without anyone logging into the OS directly.
Solna wants a single SAP-aware monitoring integration covering VM, HA, storage, and networking metrics for its SAP system. What should it configure?
A) Generic Azure Monitor workbooks built from scratch
B) Azure Monitor for SAP solutions
C) Azure Advisor
D) Azure Service Health
Solna's SAP application servers run predictable, steady-state workloads 24/7. How should it reduce compute cost without risking availability?
A) Use Azure Spot Virtual Machines for the application servers
B) Purchase an Azure Reserved Instance or Savings Plan matching the steady-state VM usage
C) Shut the VMs down every night
D) Keep pay-as-you-go pricing with no commitment
The operations team needs to start and stop a registered SAP system for planned maintenance using an Azure-native tool, without logging into the OS. What should it use?
A) Azure Center for SAP solutions
B) A generic Azure Automation runbook with no SAP awareness
C) Azure Advisor
D) The SAP LaMa connector configured entirely outside Azure
Answer Key & Rationale
1. Correct: C — Lift-shift-migrate to HANA
This strategy combines the Azure move and the HANA database conversion into one project window, matching the fixed cutover date. (A) makes no database change, so it doesn't deliver HANA before cutover. (B) splits the move and the HANA conversion into two separate phases, missing the single-window requirement. (D) is a full reimplementation, far more scope and risk than the scenario calls for.
2. Correct: A — Regional vCPU quota for the M-series family
Insufficient quota blocks the deployment outright, regardless of any other design decision, so it must be checked first. (B) affects durability and cost, not whether the deployment can happen. (C) is a cost-optimization decision made after deployment, not a blocker. (D) controls traffic flow, not capacity.
3. Correct: B — Private connectivity (VNet/ExpressRoute peering) to the RISE-managed environment
SAP RISE networking design specifically calls for private connectivity between the customer's Azure environment and the RISE-managed landscape. (A) exposes a public endpoint, contradicting the no-public-internet requirement. (C) still traverses the public internet even though it's encrypted. (D) is a web-traffic accelerator, not a private backend connectivity mechanism.
4. Correct: B — Management groups aligned to environment tiers, with Azure Policy at that level
This is the landing-zone pattern for applying governance consistently across many subscriptions while keeping environments isolated. (A) can't separate production and non-production or provide independent quota/billing boundaries. (C) enforces access control but not resource configuration rules. (D) removes any environment separation entirely.
5. Correct: B — A VM from the SAP-certified VM types list matching the SAPS rating
SAP-certified sizing is expressed and documented in SAPS specifically for that purpose. (A) ignores whether SAP has certified the VM type for this workload class. (C) burstable VMs aren't certified or suitable for sustained production SAP throughput. (D) SAP application/database tiers generally don't require GPU acceleration.
6. Correct: B — Write Accelerator enabled on Premium SSD
Write Accelerator is the managed-disk feature built specifically to cut HANA log-volume write latency on M-series VMs. (A) Standard HDD latency is far too high for active database logs. (C) Archive tier has a retrieval delay measured in hours, unusable for live transactional logs. (D) Cool tier targets infrequent access, not continuous log writes.
7. Correct: A — A proximity placement group
Proximity placement groups co-locate VMs physically to minimize inter-tier latency, which is exactly the stated requirement. (B) a Load Balancer distributes traffic; it doesn't reduce physical distance between VMs. (C) ExpressRoute connects on-premises networks to Azure, irrelevant to two VMs already in the same Azure region. (D) a larger VM size doesn't address network hops or physical placement.
8. Correct: A — The SAP Deployment Automation Framework (SDAF)
SDAF is Microsoft's purpose-built, open-source automation for repeatable SAP infrastructure deployment on Azure. (B) manual portal deployment is the opposite of repeatable automation. (C) Azure Migrate assesses and migrates general workloads; it isn't an SAP infrastructure deployment framework. (D) a generic, non-SAP Terraform module lacks the SAP-specific configuration SDAF provides.
9. Correct: B — Pacemaker with HANA System Replication and STONITH fencing
This provides automatic, database-aware failover, which is what "automatic failover for the database tier" requires. (A) an availability set alone gives VM-level redundancy but no automatic database failover logic. (C) a manual runbook is not automatic and adds unacceptable recovery time. (D) a Load Balancer doesn't track or act on database replication state.
10. Correct: B — STONITH Block Device (SBD) using iSCSI target VMs
SBD fences nodes via shared storage rather than Azure Resource Manager API calls, satisfying the no-ARM-dependency requirement directly. (A) the Azure fence agent is the option that depends on ARM APIs, which the scenario rules out. (C) no fencing risks a split-brain cluster, which HA design must prevent. (D) Load Balancer health probes detect failures but do not fence a node.
11. Correct: A — Azure Backup for SAP HANA with a policy meeting the RPO, plus periodic restore tests
A backup strategy isn't proven until a restore has actually been validated against the required RPO. (B) Site Recovery addresses regional replication/DR, not application-consistent point-in-time retention. (C) ad hoc snapshots with no policy can't guarantee a specific RPO. (D) skipping restore testing means the strategy is unverified, which is exactly what the scenario asks the team to avoid.
12. Correct: B — Azure Monitor for SAP solutions
This is the purpose-built integration covering VM, HA, storage, and networking telemetry specific to SAP systems. (A) generic workbooks require ongoing manual assembly and maintenance. (C) Azure Advisor gives point-in-time recommendations, not continuous SAP-specific telemetry. (D) Service Health reports Azure platform incidents, not SAP system health.
13. Correct: B — An Azure Reserved Instance or Savings Plan matching steady-state usage
This discounts guaranteed, predictable usage without introducing risk. (A) Spot VMs can be evicted at any time, unacceptable for a steady production SAP workload. (C) shutting VMs down nightly contradicts the stated 24/7 requirement. (D) pay-as-you-go leaves a known, committable cost saving on the table.
14. Correct: A — Azure Center for SAP solutions
ACSS provides start/stop control of a registered Virtual Instance for SAP as an Azure-native action, matching the requirement exactly. (B) a generic Automation runbook with no SAP awareness isn't the purpose-built control surface described. (C) Azure Advisor doesn't perform start/stop actions. (D) the SAP LaMa connector integrates with Azure, but the objective specifically names Azure Center for SAP solutions for this start/stop action.
FAQ
Are these questions the same difficulty as the real AZ-120 exam?
They follow the same scenario-based format Microsoft uses — a situation plus a decision, not a term-recall question — and map to the same four weighted domains. They are not reproductions of real exam items.
Why do the questions come in scenario groups instead of one at a time?
AZ-120's own exam experience frequently poses several questions against one case-study-style scenario. Practicing that format, not just isolated single-fact questions, better reflects how the real exam reads.
What if I get every question in a domain wrong?
Go back to that domain in the AZ-120 study guide and the matching section of the AZ-120 cheat sheet before attempting more questions in that area — repetition without understanding the underlying decision won't transfer to new scenario wording.
Do I need SAP Basis experience to answer these correctly?
Some working familiarity with SAP HANA, SCS, and Basis-level administration helps significantly. AZ-120 itself expects candidates to already have SAP application and database experience, not just Azure skills.
How many questions are on the actual AZ-120 exam?
Microsoft does not publish an AZ-120-specific question count on its certification page. Microsoft's general exam-duration guidance states that most certification exams contain 40–60 questions, without naming an exact figure per exam.
About These Questions
MSCertQuiz sells AZ-120 practice-exam access, and these questions were written by the same team that builds and verifies that question bank against live Microsoft Learn SAP-on-Azure documentation (checked September 7, 2026). Domain weightings above match Microsoft's official AZ-120 study guide, skills measured as of April 17, 2026.