While preparing for the AZ-104 exam, many candidates struggle with VM availability design decisions. In the enterprise world, this decision often hinges on meeting SLAs for uptime while optimizing cost and operational complexity. Let’s drill into a simulated migration scenario.
The Scenario #
Fabrikam Financial Services is modernizing their core web application infrastructure. The company plans to deploy three Azure virtual machines—named FIN-VM1, FIN-VM2, and FIN-VM3—to host their customer-facing web application, WebFinApp. Their goal is to achieve high availability so that if a single Azure datacenter (or physical failure domain) becomes unavailable, at least two VMs hosting WebFinApp remain operational.
Currently, the IT team is evaluating how to place these VMs within Azure to meet this availability SLA while adhering to company governance policies that require minimizing cross-region data egress fees and complying with regulatory data residency guidelines.
Key Requirements #
Ensure at least two out of three VMs remain available if a single Azure Availability Zone or data center failure occurs. The deployment must avoid putting all VMs into a single failure boundary.
The Options #
- A) Deploy all three VMs into a single Azure Availability Zone within one region.
- B) Deploy all three VMs into a single Azure Availability Set within the same datacenter.
- C) Deploy each VM into a separate Azure Availability Zone across the region.
- D) Deploy each VM into a separate Azure Availability Set within the same region.
Correct Answer #
C) Deploy each VM into a separate Azure Availability Zone across the region.
The Architect’s Analysis #
Correct Answer #
Option C: Each VM deployed into a separate Azure Availability Zone within the same region.
Step-by-Step Winning Logic #
Azure Availability Zones represent physically separate datacenters within an Azure region. By dispersing FIN-VM1, FIN-VM2, and FIN-VM3 into separate zones, Fabrikam Financial Services mitigates single physical site failures causing outage. This design supports the Reliability pillar of the Microsoft Well-Architected Framework (WAF). Operational Excellence is maintained because Azure manages zone isolation without heavy operational overhead.
- Availability Zones provide a 99.99% SLA for VMs.
- They reduce risk from datacenter-level failures, while adhering to regional data residency.
- Cost: Minimal extra cost, mostly related to inter-zone communication traffic which is billed modestly.
The Traps (Distractor Analysis) #
-
Option A (All VMs in single Availability Zone):
Leaves all VMs reliant on one datacenter failure domain. If that zone goes down, all VMs are unavailable—failing the requirement. -
Option B (All VMs in one Availability Set):
Availability Sets only protect from hardware and rack failures within the same datacenter but do not protect from zone or datacenter outages. A full datacenter failure would cause complete outage. -
Option D (Each VM in separate Availability Sets):
Availability Sets are not designed to span zones or datacenters; they group VMs within one datacenter. Having multiple sets in the same datacenter doesn’t protect from physical site failures.
The Architect Blueprint #
Mermaid Diagram illustrating the correct spread of VMs across Availability Zones within the same region.
Diagram Note:
User traffic is load-balanced across three VMs, each hosted in a distinct Availability Zone, ensuring zone-level fault tolerance.
The Decision Matrix #
| Option | Est. Complexity | Est. Monthly Cost | Pros | Cons |
|---|---|---|---|---|
| A | Low | Low | Simple to manage, no cross-zone latency | Single zone failure impacts all VMs |
| B | Low | Low | Protects against rack/hardware failure | No protection from zone/datacenter outage |
| C | Medium | Medium | Zone-level high availability (99.99% SLA); meets compliance | Slightly higher inter-zone data costs; more complex than sets |
| D | Medium | Medium | Some fault domain benefits within datacenter | No cross-zone protection; complex management |
Cost notes: Basic VM charges apply equally in all cases; costs rise slightly with cross-zone traffic (Option C). Availability Zones require region-specific support.
Real-World Practitioner Insight #
Exam Rule #
For the exam, always pick Availability Zones when you see the requirement “VMs must tolerate a datacenter or zone failure inside a single region.”
Real World #
In reality, many enterprises start with Availability Sets during pilot phases. As applications mature and business criticality rises, they migrate workloads into Availability Zones to meet stricter reliability SLAs, improve operational agility, and support the Microsoft CAF governance pillar for business continuity.