AZ-305 Design infrastructure solutions Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Availability Set

  Why it's wrong here

  An Availability Set logically segments VMs within a single Azure datacenter by assigning fault domains (separate racks with independent power and network) and update domains (scheduled reboot groups). This protects against localized hardware failures or patching events, but because all members reside in the same datacenter, a full datacenter outage or zone-scale disaster takes down every VM in the set. Therefore it lacks the cross-zone independence needed for datacenter-failure resilience.

• ✓

  Availability Zones

  Why this is correct

  Availability Zones are physically separate locations inside an Azure region, each with its own independent power, cooling, and networking, and each containing at least one datacenter. By placing VMs in different zones, you ensure that no single datacenter failure can bring down all instances, because the zones are designed as isolated failure domains with no shared infrastructure. This is why zone-redundant deployment is the direct answer to protecting against a full datacenter outage.

• ✗

  Virtual Machine Scale Set

  Why it's wrong here

  A Virtual Machine Scale Set automates scaling and load balancing of a fleet of identical VMs, but its default regional configuration distributes instances across fault domains within a region without a hard guarantee of placing them in separate availability zones. Unless you explicitly create a zone-redundant scale set spanning multiple zones, all instances could still land in the same datacenter, leaving them equally vulnerable to a datacenter-level failure. Thus VMSS alone does not provide the required isolation for this scenario.

• ✗

  Proximity Placement Group

  Why it's wrong here

  A Proximity Placement Group is designed to co-locate VMs in the same Azure datacenter or as close as possible, which reduces network latency for tightly coupled workloads. This actually intensifies the blast radius rather than reducing it: if that single datacenter experiences an outage, all VMs in the group fail together. It also offers no zone-level failure isolation, making it categorically unsuitable for a requirement of datacenter-failure resilience.