Which Azure VM grouping for the two 100-VM deployments?

Why the Answer Is Correct

The two rows of the exhibit are distinguished by their design keywords: Deployment 1 calls out availability zones, and Deployment 2 calls out a proximity placement group together with support for update domains. A virtual machine scale set in flexible orchestration mode treats each of the 100 instances as a standard Azure VM while still letting the set span availability zones, which is exactly the zone-resilient grouping Deployment 1 needs. Deployment 2 instead requires low-latency co-location (a proximity placement group) and staged patching across update domains, and both of those are characteristics of uniform orchestration mode rather than flexible. Scale is not the deciding factor here — both modes comfortably host 100 instances (flexible supports up to 1,000) — so the requirement keywords alone select the mode.

Why Other Candidate Answers Are Wrong

An availability set is the classic wrong pick: it distributes VMs across fault and update domains but is scoped to a single zone, so it can never satisfy Deployment 1's availability-zone requirement. Choosing uniform orchestration mode for both deployments is equally wrong, because the flexible mode's per-VM management model and zone spanning are what the first row asks for, and uniform's update-domain semantics only match the second row. Selecting a bare proximity placement group (without a scale set) is a partial answer — it addresses latency for Deployment 2 but ignores the update-domain requirement and gives no grouping mechanism for Deployment 1. Putting flexible orchestration on Deployment 2 also fails, since the PPG plus update-domain combination is the uniform-mode profile.

Community Comment Notes

mesloth condenses the scenario into two keyword pairs, quoting "Key word: proximity placement group and support for update domains" for the second deployment and noting availability zones as the keyword for the first, then mapping Deployment 1 to "VMSS Flexible" and Deployment 2 to "VMSS Uniform". That keyword-matching approach is the right mental model for this hotspot item, because the table deliberately omits any other distinguishing detail. No commenter in the thread argues for an availability set or a standalone proximity placement group, so the community reading and the documented orchestration-mode capabilities agree on the same two selections.