Lab: RCAR Analysis & Design Decision Documentation

Business Scenario: A mid-size healthcare provider (2,000 employees, 3 hospitals, 2 clinics) is migrating from legacy infrastructure to VCF. Requirements: (a) Host 400 VMs across clinical (EHR), administrative, and development workloads; (b) HIPAA compliance for all PHI workloads; (c) 99.99% availability for clinical applications (EHR system); (d) DR capability with RPO ≤ 15 min for clinical, RPO ≤ 4h for admin; (e) Self-service VM provisioning for development team (50 developers); (f) Budget: $2M capital, $400K/year operational. Document these as formal requirements with IDs (REQ-001 through REQ-006), categorized as functional, non-functional, or capacity requirements.

Identify constraints. From the scenario, extract and document at least 6 constraints: (a) Budget constraint: $2M capex, $400K/year opex — limits hardware scale and licensing tier; (b) Regulatory: HIPAA — mandates encryption at rest/in-transit, audit logging, access controls, PHI isolation; (c) Geographic: 3 hospitals + 2 clinics — determines site topology (primary DC, DR site, remote sites); (d) Timeline: assume 6-month deployment window; (e) Skills: existing team has vSphere experience but no NSX or vSAN experience — impacts operational model choice; (f) Existing infrastructure: Cisco Nexus switches must be reused for physical underlay. Format: CON-001 through CON-006 with description and design impact.

Document assumptions. For each assumption, assess: what happens if it's wrong? (a) ASM-001: 'Inter-site latency between primary DC and DR site is < 5ms RTT' — if wrong, vSAN stretched cluster is not viable, must fall back to async replication (changes DR architecture); (b) ASM-002: 'Existing Cisco Nexus switches support MTU 9000 and GENEVE encapsulation' — if wrong, switch replacement required (budget impact); (c) ASM-003: 'Development workloads do not process PHI and can be isolated in a separate workload domain without HIPAA controls' — if wrong, all domains require full HIPAA hardening (cost and complexity increase); (d) ASM-004: 'VCF 9.0 licensing includes VCF Automation for self-service catalog' — if wrong, additional licensing cost. Rank assumptions by impact: which assumption, if wrong, would require the most design revision?

Document risks. For each risk: probability (Low/Medium/High), impact (Low/Medium/High), mitigation. (a) RISK-001: 'vSAN stretched cluster inter-site link failure causes split-brain' — Prob: Low, Impact: Critical. Mitigation: dedicated link with SLA, witness at independent third site, vSAN partition handling configured for preferred site. (b) RISK-002: 'HIPAA audit finding due to incomplete micro-segmentation' — Prob: Medium, Impact: High. Mitigation: NSX DFW with zero-trust baseline, VCF Operations compliance dashboards, quarterly audit dry runs. (c) RISK-003: 'Development team creates VM sprawl exhausting capacity' — Prob: High, Impact: Medium. Mitigation: VCF Automation lease policies, resource quotas per project, monthly capacity reviews. Document at least 5 risks total.

Write formal design decisions. For each decision, use this template: Decision ID | Statement | Justification (traced to REQ/CON) | Alternatives Considered (with rejection reason) | Implications | AMPRS Impact. Write decisions for: (a) DESIGN-001: Domain topology — 'Deploy standard architecture with dedicated management domain + 2 workload domains (Clinical, Development)'. Justify from REQ-002 (HIPAA isolation) and REQ-005 (dev self-service). Alternative: consolidated architecture (rejected: CON-002 HIPAA requires workload isolation). (b) DESIGN-002: Storage architecture — 'Deploy vSAN ESA with RAID-1 for Clinical domain, RAID-5 for Development domain'. (c) DESIGN-003: DR architecture. Write at least 5 design decisions covering compute, storage, network, security, and management.

Build a Requirements Traceability Matrix (RTM). Create a spreadsheet with columns: Requirement ID | Requirement Description | Design Decision(s) | Validation Method | Status. Map every requirement to at least one design decision. Map every design decision to a validation method: 'functional test', 'performance benchmark', 'failover test', 'compliance audit', or 'capacity review'. Verify completeness: are there any requirements without design decisions? Any design decisions without backing requirements?

Conduct AMPRS analysis for the complete design. For each dimension: (a) Availability: list all HA mechanisms deployed and their coverage — vSphere HA (VM restart), vSAN stretched cluster (site failover), NSX Edge HA (network path). Identify single points of failure that remain. (b) Manageability: assess operational complexity — how many consoles, how many skills required, what can be automated. (c) Performance: identify potential bottlenecks — vSAN write latency with RAID-1 across sites, NSX overlay overhead. (d) Recoverability: map RPO/RTO to protection mechanisms per workload tier. (e) Security: enumerate all security controls by layer (network, compute, storage, identity, compliance).

Simulate a VCDX defense challenge. For each design decision, prepare for these question types: (1) 'Why not X?' — prepare alternative analysis. Example: 'Why vSAN stretched cluster instead of SRM?' Answer: 'REQ-003 requires 99.99% — SRM RTO of 30+ minutes cannot meet this SLA. Stretched cluster provides automatic failover with RTO < 5 minutes.' (2) 'What if assumption Y is wrong?' — walk through design revision. (3) 'What happens when Z fails?' — trace failure through AMPRS dimensions. Practice defending DESIGN-001 through DESIGN-005 against at least 2 challenge questions each.

Review and refine. (a) Check every design decision for completeness — does it have justification, alternatives, implications, and AMPRS impact? (b) Check RTM coverage — are all requirements addressed? (c) Check assumption ranking — is the highest-impact assumption clearly identified? (d) Check risk mitigations — is every High-impact risk mitigated? (e) Self-critique: identify 2 weaknesses in your own design and document how you would address them with more budget/time.