Skip to main content
In this lesson we cover how to plan virtual machines (VMs) in Azure and how the cloud shared responsibility model affects securing and operating those VMs. Proper planning helps balance performance, availability, security, and cost. The shared responsibility model in cloud computing defines which security and operational tasks are handled by the cloud provider and which remain the customer’s responsibility. The split depends on the service model (IaaS, PaaS, SaaS). For IaaS (virtual machines), the cloud provider (Microsoft Azure) manages the physical infrastructure — data center security, physical hosts, the underlying network, and the virtualization layer — while the customer is responsible for everything inside the VM: guest OS, updates and patches, network and access configuration, endpoint protection, applications, and data. As you move up the stack to PaaS and SaaS, Microsoft assumes more responsibilities (runtime, middleware, and sometimes the OS for PaaS; almost the entire stack for SaaS). Regardless of model, customers always retain responsibility for their data: classification, ownership, encryption, backups, and access controls. Understanding this model is essential when planning VMs so that security, compliance, and operational tasks are assigned and implemented correctly.
A diagram titled "Shared Responsibility Model" showing a table of responsibilities (e.g., information & data, devices, accounts, OS, physical hosts) across service types SaaS, PaaS, IaaS and On‑Premises. Colored dots indicate whether each responsibility is held by the customer (teal) or Microsoft (gray) for each service model.
Never assume all security is handled by the cloud provider — for IaaS VMs you remain responsible for the guest OS, installed software, patching, endpoint protection, and data governance. Confirm responsibilities before deploying.

​
Key planning areas for Azure VMs

When planning VMs in Azure, focus on the areas below. These will guide design decisions to meet performance, availability, and cost objectives.
  • Networking
    • Design the virtual network (VNet) and subnets for current and future scale.
    • Avoid overlapping IP ranges with on-premises networks or peered VNets to prevent routing conflicts.
    • Choose connectivity patterns (VPN, ExpressRoute, VNet Peering) and design routing and security (Network Security Groups, Azure Firewall, user-defined routes).
  • Naming
    • Create a consistent naming convention that includes environment (dev/test/prod), role (web/db), service, and region.
    • Consistent names simplify automation, billing, policy enforcement, and troubleshooting.
  • Location (Region)
    • Verify that required VM sizes and Azure services are available in your chosen region.
    • Consider latency-sensitive user proximity and data residency or compliance requirements.
    • Be aware that pricing and capacity can vary by region.
  • Pricing and cost optimization
    • Understand pricing models and select the best fit:
      • Pay-as-you-go for flexibility.
      • Reserved Instances (or Azure Reserved VM Instances) for steady-state workloads.
      • Spot VMs for interruptible, non-critical tasks with deep discounts.
    • Use Azure Hybrid Benefit to apply eligible on-premises Windows Server or SQL Server licenses to reduce costs.
    • Monitor and control costs with Azure Cost Management and budgeting tools.
Avoid overlapping IP address ranges with on-premises networks or other VNets you intend to connect. Overlaps can prevent successful routing and complicate network design.

​
VM planning checklist

The table below summarizes practical actions to take when planning VM deployments.

​
Additional resources

A slide titled "Virtual Machine Planning" showing four colorful panels labeled Networking, Naming, Location, and Pricing with icons and bullet points. It summarizes key planning considerations (address spaces, naming conventions, region/size choices, and pricing options) for deploying VMs.
On that note, let’s take a look at the VM options, sizing, and configuration choices you should evaluate next.

Watch Video