Public sample report
See what the workload evidence can tell you.
Example workflow — not a customer case study.
This public sample uses fictional VMware, Proxmox, KVM and Kubernetes data to show the evidence format: allocation, observed signals, missing data, risks and decision notes. The rows are separate workloads, so this page does not invent a false before/after pair.
Demonstration data only. No customer infrastructure or live system is connected.
Infrastructure overview
Fictional workloads: 4
Observation window: Not available
N/A means unavailable, not zero. Field presence does not establish representative time coverage.
Allocation and observed utilisation
| Workload | Platform | Allocated vCPU | Allocated RAM (GiB) | CPU snapshot | RAM snapshot | Signal completeness |
|---|---|---|---|---|---|---|
| prod-payments-api-01 | VMware | 4 | 8 | 18.4% | 61.2% | Partial |
| prod-erp-db | Proxmox | 8 | 16 | 34% | 75% | Partial |
| prod-web-02 | KVM/libvirt | 2 | 4 | N/A | 48.5% | Limited |
| prod/payments-api-7f8cc9d78b-x42pz | Kubernetes | 0.6 | 0.875 | N/A | N/A | Limited |
Completeness reflects the report’s available CPU/RAM fields. It is not confidence in a final sizing decision. Kubernetes allocations are resource requests; they are not measured consumption.
Coverage, risks and prerequisites
The fixtures contain no observation period or time series. CPU/RAM p95, p99, peak behaviour and business-cycle coverage are unavailable. A short measurement never guarantees a suitable target size.
VMware guest-memory observations and Proxmox memory accounting have different meanings. VMware datastore pressure is also different from workload disk utilisation. A shared percentage label does not make these measurements comparable.
Before any pilot, confirm application ownership, dependencies, backups, acceptance criteria, a maintenance window and a rollback plan with the team responsible for the change.
Disk I/O, network throughput, energy, application performance and anomaly history: N/A. No savings or energy reduction is calculated.
Decision: further review required
This evidence can frame the next observation and technical review. It does not approve a target configuration or authorise a production change.
How before/after validation is delivered today
For a real project, InfraWatts captures the agreed workload scope before the change, repeats the observation after the move and the partner reviews workload correspondence before the delta report is finalised. The cross-platform matching step is not yet fully automatic.
Review a before/after validation scope for your migration.
Share the source platform, target, approximate workload count and migration window. We define the pre-change baseline, post-change observation and matching method before production processing starts.
Review a migration case