Azure Migrate Datacenter Discovery and Assessment

The free call is a conversation: your situation, your drivers, and an honest steer on direction — including whether this assessment is worth buying at all. This engagement is two weeks of measurement: an appliance in your environment, observed utilization, mapped dependencies, and a costed model you can put in front of a CFO. Start with the call; it is free, and it exists to stop you spending $1,950 you might not need to spend. This page is the step after it, never a replacement for it.

Because right-sizing from a snapshot is guessing with confidence. A server sampled at 9 a.m. Tuesday looks nothing like the same server during Sunday-night batch runs. Two weeks of collection captures real weekly cycles, so recommendations are based on observed utilization patterns rather than a single flattering or damning moment — and it is the difference between a cost model a finance team can trust and one they will rightly pick apart. If your estate has month-end or seasonal spikes beyond the window, we say so and size with stated headroom.

It is a Microsoft-provided discovery appliance running in your environment that inventories servers and polls performance and connection data, sending assessment metadata outbound over HTTPS to your Azure Migrate project. Discovery is read-only: it does not change server configuration, install anything on VMware guests, or sit in any data path. The polling load is light by design. We deploy it, verify it is healthy, and either remove it cleanly at the end or leave it for your continued use — your call.

For VMware estates, no — discovery and dependency analysis are agentless through vCenter APIs. For Hyper-V and physical servers, dependency data is gathered over standard remote management — WinRM for Windows, SSH for Linux — still without installing persistent agents on the workloads. The exact mechanism per platform is confirmed against Microsoft's current documentation during scoping, and every credential we use is agreed in writing, read-only wherever the platform allows.

Microsoft provides the Azure Migrate discovery and assessment tooling at no additional charge in your subscription — 'free' here is Microsoft's tooling, not our work; the $1,950 buys the engineering: deployment, a monitored collection window, dependency corroboration, right-sizing analysis, the cost model, and the wave plan. Trivial Azure charges can arise from the project's supporting resources, and anything that runs in your subscription is billed by Microsoft to you — stated so the invoice contains no mysteries.

As accurate as evidence and stated assumptions allow, and no more. Sizes come from observed utilization with headroom, not nameplate specs; run-rates are computed at Microsoft's published pricing on the delivery date across pay-as-you-go, reservation, and savings-plan options; Azure Hybrid Benefit is applied only where your license position actually qualifies. Every assumption is written down so you can challenge it. What it is not is a guarantee of your future bill — Azure pricing moves, consumption behavior moves, and anyone promising your exact bill a year out is decorating.

It is the observed record of which servers actually communicate — not which ones the documentation says communicate. Migrations fail at the seams: an application server moves to Azure while the license server it silently depends on stays behind, and Monday morning is a post-mortem. Dependency data groups interdependent servers into move groups that migrate together, and we corroborate the groupings with your application owners because the tool sees connections while your people know what they mean. It is the single best predictor of a quiet cutover.

Then that is what the report says, in plain language, with reasons per server. Real estates almost always split: workloads that move cleanly now, workloads that need remediation first, servers whose job should be retired or bought as SaaS, and a few that genuinely belong on-premises. You are paying for the truth of that split, not for a slide deck that concludes 'migrate everything, with us.' A do-not-migrate verdict costs us migration revenue and still goes in the report — that is the point of charging for the assessment instead of giving it away as a sales motion.

Yes at the discovery level: SQL Server instances and databases are discovered and inventoried, and web applications surface in the estate picture, all feeding the wave plan. Where a serious database estate emerges, we recommend the dedicated SQL Server to Azure migration assessment for engine-level analysis — target options from Azure VM to managed services deserve more than a VM-sizing pass. The wave plan tells you when that deeper step is warranted and when it would be over-engineering.

Sometimes. Microsoft periodically operates partner-delivered funding and incentive programs for qualifying Azure migrations; eligibility, scope, and amounts are Microsoft's decision, change by fiscal year, and are never guaranteed. During the assessment we will tell you honestly whether your estate plausibly fits the current programs and help pursue them if it does — but the cost model never depends on funding materializing. Plans built on incentives that then lapse are how migrations stall halfway.

You decide, from evidence. Most clients take the wave plan and convert wave one into fixed per-workload quotes — Windows Server, SQL Server, web apps — each a separately priced engagement with its own written scope. Some pause at prerequisites, typically a landing zone, and do that first, which is often the right call. Some conclude the numbers favor staying put for now, protect the estate with replication instead, and revisit later. All three are legitimate outcomes, and the walkthrough ends with a recommendation, not a close.

Yes. The inventory, dependency map, cost model, and wave plan are working files delivered to you, and the Azure Migrate project lives in your subscription — there is nothing proprietary holding the findings hostage. We would obviously like the migration work, and we price it in fixed, per-workload written quotes so comparison is easy; but the assessment is designed to stand on its own, because an assessment that only makes sense if you hire its author is an advertisement.