Sign in to practice all 74 questions with progress tracking and AI explanations.
1. A startup wants to move fast without managing servers and only pay when code runs. Which compute abstraction best fits their need?
- A.Serverless functions provide event-driven execution with automatic scaling and billing only for actual invocations and runtime duration.✓ Correct
- B.IaaS virtual machines provide full OS control but require patching, capacity planning, and continuous instance billing regardless of use.
- C.PaaS web apps eliminate coding needs entirely and guarantee zero operational tasks across all workloads by default.
- D.Bare metal hosts deliver maximum performance and isolation with built-in autoscaling and per-request metering out of the box.
Explanation
Serverless charges per execution and abstracts servers. IaaS still manages VMs; PaaS reduces ops but not per-invocation; bare metal lacks serverless autoscaling and metering.
2. A legacy .NET app with Windows services and COM components must migrate quickly with minimal code changes. Which approach is most appropriate?
- A.Rehost to Azure Virtual Machines to preserve OS-level dependencies and minimize refactoring while meeting tight timelines.✓ Correct
- B.Rewrite the application as microservices on AKS to improve scalability and fully remove all legacy dependencies immediately.
- C.Package the app in Azure Container Apps to avoid any OS considerations and guarantee compatibility with COM components.
- D.Move to Azure Functions to eliminate process-level services and schedule triggered executions without persistent runtimes.
Explanation
Lift-and-shift to VMs keeps Windows service and COM dependencies. AKS/Functions/Container Apps require refactor or lack support for such OS-level features.
3. A high-throughput OLTP database needs large memory for caching and balanced CPU. Which Azure VM family is the best starting point?
- A.Memory Optimized VMs prioritize higher RAM per vCPU, benefiting databases that rely on large buffer caches and in-memory operations.✓ Correct
- B.Compute Optimized VMs provide high CPU per memory ratios that are ideal for stateless batch processors with minimal caching needs.
- C.GPU VMs include specialized cores for graphics and training, making them optimal for transactional database workloads at scale.
- D.Storage Optimized VMs reduce memory to increase NIC bandwidth exclusively, solving all IO bottlenecks for OLTP databases.
Explanation
Memory Optimized fits large caches for OLTP. Compute Optimized suits CPU-bound tasks; GPU targets graphics/AI; Storage Optimized focus on disk throughput, not memory size.
4. You must provide cost-effective Windows 11 Enterprise session host access for 200 users with intermittent usage. What design choice helps optimize cost and scale?
- A.Use pooled multisession host pools with autoscaling to right-size capacity during peak and scale-in aggressively when users disconnect.✓ Correct
- B.Use personal single-session VMs per user to isolate workloads and run 24x7 to avoid any reconnection delays or cold starts.
- C.Assign all users to a single large VM without autoscaling to reduce management complexity and guarantee consistent performance.
- D.Disable FSLogix profiles to reduce storage costs and rely on local profiles that reset during every user logoff.
Explanation
Pooled multisession with autoscaling reduces cost for intermittent users. Personal VMs and fixed capacity raise costs; disabling FSLogix harms user experience and persistence.
5. A team needs to run a short-lived containerized job nightly without managing nodes or clusters. Which Azure service is the simplest fit?
- A.Azure Container Instances run containers on demand without cluster management and bill only for the resources used during execution.✓ Correct
- B.Azure Kubernetes Service provides full cluster control and requires node scaling policies even for small scheduled batch jobs.
- C.Azure Virtual Machines provide the lowest cost for tiny jobs because instances are free when stopped and incur no storage charges.
- D.Azure Container Apps requires building a complex microservices mesh before any single container can execute a scheduled job.
Explanation
ACI is serverless for containers and ideal for short jobs. AKS and VMs require managing compute; Container Apps does not mandate complex meshes for one-off jobs.
Practice all 74+ questions in this domain
Start free practice →