LevelUpWith

Explore / Microsoft Azure Solutions Architect (AZ-305) Exam Prep / Practice questions

Microsoft Azure Solutions Architect (AZ-305) practice questions

10 questions from the bank behind this course's timed mock. Free to read and to answer, no account needed, and each has its answer and an explanation. They belong to Microsoft Azure Solutions Architect (AZ-305) Exam Prep.

Answer them one at a time →Read lesson 1 free

  1. Question 1

    You are an Azure architect for an insurance company. The requirements for a claims portal say it must cost less to host and need the least administrative effort. A senior engineer recommends running it on virtual machines because they give full control over the operating system and can meet the performance target. Using the architect's decision method, what should you do with the virtual machine option?

    1. Keep it and add a new requirement for operating system control so that it qualifies
    2. Recommend it, because it meets the performance target, which matters most on AZ-305
    3. Recommend it, because full control over the operating system is the most flexible design
    4. Rule it out when comparing trade-offs, because it fails the stated least-administrative-effort constraint
    Show the answer

    Answer: D. Rule it out when comparing trade-offs, because it fails the stated least-administrative-effort constraint

    During the trade-offs step you rule out any option that fails a hard constraint, even if it is a good service in general. Virtual machines need patching and management, so they fail 'least administrative effort'. Recommending them for flexibility ignores the stated requirement, and the requirements come from the scenario, not from the architect.

  2. Question 2

    You are the solutions architect for an ISV that is building a multi-tenant SaaS application. Customer organisations will consent to the app in their own Microsoft Entra tenants so it can call Microsoft Graph for them. The app is hosted by a third-party cloud provider, not in Azure. Which identity should you recommend for the application?

    1. A system-assigned managed identity
    2. A service principal created from an app registration
    3. B2B collaboration guest accounts for each customer organisation
    4. A user-assigned managed identity shared by all customer tenants
    Show the answer

    Answer: B. A service principal created from an app registration

    A multi-tenant SaaS app that runs outside Azure needs an app registration and its service principal. Managed identities work only with Azure resources, so neither the system-assigned nor the user-assigned option fits here.

  3. Question 3

    You are a solutions architect at Fabrikam. In Cost analysis, the finance team groups costs by the env tag and sees separate lines for 'Production', 'production' and 'PRODUCTION'. The tag key is spelled several ways too: 'Env', 'env' and 'ENV'. You need one cost line per environment, with the least redesign. What should you recommend?

    1. Standardise the tag values to a single agreed casing, such as 'prod', and publish it in the tagging taxonomy, because tag values are case-sensitive.
    2. Move the env tag from the resources to the management group so that every subscription reports one value.
    3. Turn on Cost Management tag inheritance so that the resource group's env tag replaces the resource's own tag in usage records.
    4. Standardise the tag key to a single casing, such as lowercase 'env', because tag names are case-sensitive and cause the split.
    Show the answer

    Answer: A. Standardise the tag values to a single agreed casing, such as 'prod', and publish it in the tagging taxonomy, because tag values are case-sensitive.

    Tag values are case-sensitive, so 'Production' and 'production' are reported as separate groups. Fixing the allowed values and their casing removes the split. Changing the key's casing would not help, because tag names are case-insensitive and the different spellings of the key are already treated as one tag.

  4. Question 4

    You are the solutions architect for a manufacturer that is moving a 600 GB SQL Server database to Azure SQL Database. The company has SQL Server licences with Software Assurance that it wants to reuse, and it plans to commit to a one-year reservation to lower compute cost. The database's storage needs will grow faster than its compute needs. Which purchasing model should you recommend?

    1. The DTU-based purchasing model in the Premium tier
    2. The DTU-based purchasing model in the Standard tier, with storage included in the bundle
    3. The vCore-based purchasing model with Azure Hybrid Benefit and reserved capacity
    4. The DTU-based purchasing model in an elastic pool using eDTUs
    Show the answer

    Answer: C. The vCore-based purchasing model with Azure Hybrid Benefit and reserved capacity

    Only the vCore model offers Azure Hybrid Benefit for existing licences and reserved capacity, and it lets you scale compute and storage separately. DTU Premium is a real, high-performance tier, but it sells compute and storage as a preconfigured bundle and does not support reservations or licence reuse.

  5. Question 5

    You are designing an Azure Cosmos DB for NoSQL container for a multi-tenant SaaS application. Most queries filter on tenant and then on user, for example "get the sessions for user U in tenant T". Most tenants are small, but one enterprise tenant already holds more than 30 GB of data and is still growing. You need to keep the main queries scoped to as few partitions as possible and avoid hitting the storage limit of a logical partition. Which partition key strategy should you recommend?

    1. A partition key of /id, because it has the highest cardinality
    2. A single partition key of /TenantId
    3. A hierarchical partition key of TenantId/UserId/SessionId
    4. A partition key of /tenantTier, which groups tenants by subscription plan
    Show the answer

    Answer: C. A hierarchical partition key of TenantId/UserId/SessionId

    Hierarchical partition keys support up to three levels, so one large tenant can grow past the 20 GB logical-partition limit while tenant- and user-scoped queries stay targeted. /TenantId alone puts the whole enterprise tenant in one logical partition, which can't hold more than 20 GB.

  6. Question 6

    You are a solutions architect for a retailer. One regional deployment of the checkout workload has a composite availability of 99.84%. You deploy it to two regions in parallel; either region can serve all traffic. All requests reach the regions through Azure Front Door, which you use for this calculation at 99.99% availability. What is the approximate composite availability of the full design?

    1. 99.9997%
    2. 99.84%
    3. 99.68%
    4. 99.99%
    Show the answer

    Answer: D. 99.99%

    Two regions in parallel give 1 − (0.0016)² ≈ 99.9997%. Front Door is in series with both regions, so you multiply by 99.99%, which gives about 99.99%. Choosing 99.9997% means you forgot that every request passes through the global router.

  7. Question 7

    Contoso runs a legacy licensing server on one Azure VM. The vendor's licence allows only one instance, so you cannot add more VMs. The OS disk and data disk both use Standard SSD. Management wants the highest VM connectivity SLA possible for this single VM at the lowest added complexity. What should you recommend?

    1. Convert only the OS disk to Premium SSD and keep the data disk on Standard SSD
    2. Redeploy the VM as a zonal VM in availability zone 1
    3. Place the VM in an availability set with three fault domains
    4. Convert both the OS disk and the data disk to Premium SSD
    Show the answer

    Answer: D. Convert both the OS disk and the data disk to Premium SSD

    A single VM gets the 99.9% SLA only when all of its disks are Premium SSD or Ultra Disk. With Standard SSD the SLA is 99.5%, so converting only the OS disk is not enough. An availability set or a zone raises the SLA only when two or more VMs are deployed, which the licence does not allow.

  8. Question 8

    You are planning a migration for a company that is moving two applications to Azure. App A is a stateless web API written for a runtime that Azure App Service supports, and it has no OS dependencies. App B is a vendor application that is installed with a setup wizard, needs a specific Windows Server version and changes the registry during installation. The company wants as little operational overhead as possible. What should you recommend?

    1. Deploy both applications to Azure App Service
    2. Deploy App A to Azure App Service and App B to Azure virtual machines
    3. Deploy App A to Azure virtual machines and App B to Azure App Service
    4. Deploy both applications to Azure virtual machines in a Flexible scale set
    Show the answer

    Answer: B. Deploy App A to Azure App Service and App B to Azure virtual machines

    App B's setup-wizard installation, OS version dependency and registry changes are legacy and OS-control constraints that justify a VM, while nothing rules out PaaS for App A, so the managed service is the better choice. Putting both on VMs adds operational work for App A without a reason.

  9. Question 9

    You are designing API governance for a public weather API. Anonymous consumers call it without subscription keys. The API owner requires that each client IP address can make no more than 60 calls per minute, so that a single caller cannot overload the back end with short bursts. The instance runs in the classic Standard tier. Which policy should you recommend?

    1. Limit call rate by key (rate-limit-by-key), using the caller's IP address as the counter key
    2. Set usage quota by key (quota-by-key), using the caller's IP address as the counter key
    3. Limit call rate by subscription (rate-limit) set to 60 calls per 60 seconds
    4. Restrict caller IPs (ip-filter), allowing only known address ranges
    Show the answer

    Answer: A. Limit call rate by key (rate-limit-by-key), using the caller's IP address as the counter key

    rate-limit-by-key can throttle short bursts on any expression, including the client IP, and it does not need a subscription key. The rate-limit policy counts calls per subscription, so it cannot throttle anonymous callers, and quota-by-key caps total volume over a longer renewal period instead of absorbing per-minute spikes.

  10. Question 10

    A retailer is migrating a 3 TB SQL Server 2019 database to Azure. A third-party monitoring and backup agent must stay installed on the database host, and the database administrators need full access to the operating system. The business allows only a few minutes of downtime at cutover. What should you recommend?

    1. Migrate to SQL Server on Azure Virtual Machines by using an online migration in Azure Database Migration Service
    2. Migrate to SQL Server on Azure Virtual Machines by using an offline migration in Azure Database Migration Service
    3. Migrate to Azure SQL Managed Instance by using the Managed Instance link
    4. Migrate to Azure SQL Database by using an online migration in Azure Database Migration Service
    Show the answer

    Answer: A. Migrate to SQL Server on Azure Virtual Machines by using an online migration in Azure Database Migration Service

    The software on the host and the need for OS access both require SQL Server on Azure VMs, and DMS supports online migration to that target, so downtime is limited to the cutover. Managed Instance with the Managed Instance link would give minimal downtime, but it is PaaS and gives no OS access and no way to install software on the host.

The full timed mock

50 questions in 120 minutes, with a pass mark of 70%. Each paper is drawn from a bank of questions written for the mock alone, and these 10 are left out of it. Full access opens it, with every level of the course.

See the course →See the plans

LevelUpWith is independent of the organisation that sets this exam. These questions were written for this course; they are not taken from the exam.