A company has a number of vSphere clusters where it runs business-critical workloads. Company policy states that to ensure adequate resources, the clusters should never have greater than a 4:1…

Question

Options

• AAllocation
• BDemand
• CLimit
• DReservation

• A
  70% (26)
• B
  8% (3)
• C
  5% (2)
• D
  16% (6)

Explanation

Allocation is correct because this capacity model measures what resources have been provisioned to VMs (configured vCPUs and vRAM), not what they're currently consuming. Since the company's policy enforces specific virtual-to-physical ratios (4:1 CPU, 1:1 memory), vROps needs to track everything that has been promised to VMs regardless of actual usage - that's exactly what the Allocation model does.

Demand is wrong because it measures real-time or historical resource consumption. A cluster could technically stay under the demand threshold while still violating the allocation ratio, meaning over-provisioned VMs would go undetected until they all spike simultaneously.

Limit tracks resource limits configured on individual VMs; if no limits are set, it assumes unlimited capacity - making it useless for enforcing an organization-wide ratio policy.

Reservation only accounts for resources explicitly reserved via vSphere guarantees. Since most VMs don't have reservations configured, this would massively undercount actual allocated resources and give a false picture of available capacity.

Memory tip: Think "A for Allocated, A for Agreement" - when your policy is about a contractual ratio (what you've agreed to give VMs), use Allocation. Demand is for "what's actually being used," which is a different question entirely.

No community discussion yet for this question.