3V0-21.23 · Question #119
An architect is designing a new vSphere-based solution for a customer. During a requirement gathering workshop, the following information is provided: - The solution must have a primary and…
The correct answer is A. 12. Option A (12 hosts) is correct because it satisfies all constraints across both sites when you work through each cluster's requirements, driven by the RAM bottleneck and the disaster recovery scenario. The math breaks down like this: Primary site (production only): Half of…
Question
An architect is designing a new vSphere-based solution for a customer. During a requirement gathering workshop, the following information is provided:
- The solution must have a primary and secondary site.
- The solution must support a maximum of 1,000 concurrent workloads.
The profile of the workloads are as follows:
- Production Workloads
-- 300 x Small: 1 vCPU, 2 GB RAM -- 400 x Medium: 2 vCPU, 6 GB RAM -- 100 x Large: 4 vCPU, 8 GB RAM
- Development Workloads
-- 200 x Small: 1 vCPU, 2 GB RAM The corporate security policy states that, during normal operations, production workloads must be physically segregated from development workloads. All production workloads are split evenly across the primary and secondary site. All development workloads run only within the secondary site. In the event of a disaster affecting workloads in the primary site, the secondary site must be capable of running all production and development workloads. The vCPU to physical core ratio should be a maximum of 10:1 for production workloads and 20:1 for development workloads. The solution should provide a minimum of N + 1 resiliency at each component level. The target physical host hardware platform has already been defined by the company's hardware standards and therefore each host has the following configuration:
-- 2 x 24 physical cores -- 768 GB RAM -- 2 x 100 GB SSD drives -- 6 x 10 GbE network cards What is the minimum number of hosts required to meet the requirements?
Options
- A12
- B14
- C8
- D10
How the community answered
(28 responses)- A64% (18)
- B11% (3)
- C21% (6)
- D4% (1)
Explanation
Option A (12 hosts) is correct because it satisfies all constraints across both sites when you work through each cluster's requirements, driven by the RAM bottleneck and the disaster recovery scenario.
The math breaks down like this:
- Primary site (production only): Half of production = 750 vCPU / 1,900 GB RAM. RAM is the binding constraint: ceiling(1,900/768) = 3 active hosts needed, so N+1 = 4 hosts.
- Secondary site - production cluster: During a disaster, it must absorb all production (1,500 vCPU / 3,800 GB RAM). RAM again drives: ceiling(3,800/768) = 5 active hosts, so N+1 = 6 hosts. (Note: 5 × 768 = 3,840 GB - just barely enough.)
- Secondary site - development cluster (physically segregated): 200 vCPU / 400 GB RAM at 20:1 ratio fits on 1 active host, so N+1 = 2 hosts.
- Total: 4 + 6 + 2 = 12 hosts.
Why distractors fail:
- D (10) likely undercounts by sizing the secondary production cluster for normal operations (4 hosts) rather than the disaster scenario (6 hosts).
- C (8) likely ignores N+1 resiliency entirely and uses raw minimums - perhaps 4 + 2 + 2.
- B (14) likely overcounts by applying N+1 redundantly (e.g., adding a spare per site on top of already-inflated numbers).
Memory tip: Always identify the binding constraint for each cluster (vCPU ratio vs. RAM), then apply N+1 after sizing - and remember that disaster recovery sizing, not normal operations, determines how many hosts the secondary site needs.
Topics
Community Discussion
No community discussion yet for this question.