Solution Design, Sizing, and Pre-Sales Discovery
What this domain tests
Discovery, requirements, sizing methodology, reserve, N+1, performance, and TCO.
Core rules and concepts
- Discover business requirements first: workloads, SLAs, RPO/RTO, growth, compliance, and constraints.
- Size usable capacity after RF, metadata, snapshot reserve, and rebuild headroom. Never size on raw capacity.
- Include N+1 node failure tolerance and growth runway.
- Performance sizing uses IOPS, throughput, latency, and working set. All-flash versus hybrid changes the answer.
- Use Nutanix Sizer and Collector for data-driven designs. Validate high-risk designs with a proof of concept.
- Licensing and support affect TCO and must be included.
Worked example
100 VMs need 2 TB usable, 10k IOPS, 20% growth, RF2, and N+1. Raw capacity starts at 2 TB times 2 for RF, or 4 TB. Add 25% overhead, giving 5 TB. Add N+1, giving about 6.25 TB. Add 20% growth, giving about 7.5 TB. Choose a node count that meets both capacity and IOPS. If four nodes provide 8 TB usable but only 8k IOPS, the design still fails performance.
Traps
Do not size raw capacity. Do not ignore reserve or N+1. Do not let a vendor SKU drive requirements. Do not promise performance without a workload profile. Do not forget backup and DR capacity.