TL;DR
Pick Azure migration tools by workload: Azure Migrate for servers, Database Migration Service for SQL Server and Data Box for bulk data. Start with Azure Migrate, the free hub that discovers, assesses and then moves VMware, Hyper-V, physical and other-cloud servers. The tooling itself is free, and server replication bills only after 180 days per machine. Storage Mover or AzCopy move file shares online, but 200 TB over 1 Gbps takes about 26 days, so ship that on Data Box. Web apps go to App Service, in bulk through Azure Migrate or one site at a time with the Migration Assistant. Data Migration Assistant retired on July 16, 2025, so since then SQL Server assessments run in Azure Migrate and the moves run in DMS.
Key Takeaways
- Azure Migrate is the hub: it discovers, assesses and migrates servers, and it also assesses SQL Server instances and web apps.
- Azure Database Migration Service moves SQL Server to Azure SQL Database, SQL Managed Instance or Azure VMs, while Azure Migrate handles the assessment.
- Data Box, Storage Mover and AzCopy cover files and bulk data, and the right one depends on volume and bandwidth as well as the deadline.
- Site Recovery is a protection and replication service, while Resource Mover works only between Azure regions.
- Tool charges are small compared to dual running, egress and idle capacity, and Flexera puts wasted cloud spend at 29%.
- Native tools stop at the server, so pipelines, semantic models and reports on those servers need separate conversion work.
Watch on YouTube
Azure Migration Strategy: Hybrid Cloud, Cost Optimization & AI
Kanerika walks through Azure migration strategy across hybrid cloud, cost optimization and AI, so you can see the decisions that sit around the tool choices in this guide.
Four Workloads and a Friday Deadline
Picture an IT lead with a migration plan due on Friday. The estate holds 400 VMware virtual machines, 30 SQL Server instances, 200 TB of file shares and a handful of .NET web apps. The data center lease is up for renewal, so the CFO wants a tool list, a cost line and a cutover order.
A first search returns a dozen Microsoft product names. Several are retired and a few are still in preview. Some move servers and some move databases, while others only copy files.
One of them, Site Recovery, appears on almost every list even though its main job is disaster recovery. That mix-up alone can send a project down the wrong path for weeks before anyone notices.
This guide sorts Azure migration tools by the workload each one moves. It also covers what each tool costs, what changed in 2025 and 2026, and where native tooling stops. Later on, we return to the Friday estate and build its plan step by step, with the transfer math included.
What Are Azure Migration Tools?
Azure migration tools are the Microsoft services that find what you run on premises or in another cloud and move it into Azure. They fall into two groups. Azure Migrate is the hub, while a set of specialist tools handles databases, bulk data and files.
The hub is where nearly every project starts, because discovery comes before any move. According to Microsoft’s Azure Migrate overview, the service discovers servers, databases and web apps, then checks whether each one is ready for Azure. It also right-sizes the target and builds a business case that compares on-premises and Azure costs. Servers it migrates itself, while large offline transfers go to Azure Data Box.
One phrase on the pricing page still causes real confusion. Microsoft says Azure Migrate is available at no additional charge, but that covers the tooling only. The virtual machines, storage and databases you land in Azure bill once they start running. Third-party tools used inside Azure Migrate may also charge separately.
Tools carry out a plan; they do not write one. Deciding which workloads to retire, rehost or rebuild belongs to your cloud transformation strategy. That call should come before anyone installs an appliance. Once that is settled, the toolkit breaks into four layers:
- Discovery and assessment: Azure Migrate, including dependency analysis and the business case.
- Server and app migration: Azure Migrate server migration and App Service Migration Assistant.
- Database migration: Azure Database Migration Service, once Azure Migrate has assessed the databases.
- Data and file transfer: Azure Data Box, Azure Storage Mover and AzCopy.
Each layer has a clear owner in a well-run program. Our roundup of data migration tools covers the third-party options around this core. Infrastructure teams usually run the first two, while database teams own the third and storage teams take the fourth.
The Azure Migration Tools at a Glance
Each Microsoft migration tool has a narrow job, so see the whole set on one page first. Read the table as a routing chart. Find the thing you need to move, then read across to the tool, the transfer mode and its 2026 status.
| Tool | Moves what | Online or offline | Tool cost | 2026 status |
|---|
| Azure Migrate | VMware, Hyper-V, physical and other-cloud servers; assesses SQL Server and web apps | Online replication | No additional charge; server migration free for 180 days per machine, then billed monthly | Active; Copilot migration agent and Auto Wave Planning in preview |
| Azure Database Migration Service | SQL Server to Azure SQL Database, SQL Managed Instance or SQL Server on Azure VMs | Online and offline (SQL Database offline only) | Standard tier free for offline moves; Premium (online) billed per vCore after 183 free days | Active; SQL Server targets only |
| App Service Migration Assistant | .NET web apps from Windows to Azure App Service | Online | You pay for the App Service plan | Standalone tool |
| Azure Data Box | Bulk data on 120 TB or 525 TB devices | Offline, shipped device | Order and shipping fees | Data Box Heavy retired; 80 TB device retiring by region |
| Azure Storage Mover | SMB shares to Azure Files or Blob, NFS shares to Azure Files, and AWS S3 to Blob | Online, managed | Service free; you pay for target storage and transactions | AWS FSx SMB and Google Cloud Storage sources in preview |
| AzCopy | Files and blobs to, from or between storage accounts | Online, command line | You pay for target storage | v10 is the current release |
| Azure Site Recovery | Replicated machines for disaster recovery | Online replication | Free for 31 days per protected instance, then billed per instance | Active |
| Azure Resource Mover | Azure resources between Azure regions | Online | You pay for the resources in the target region | Active |
Azure Migrate: Discovery, Assessment, Business Case and Server Migration
Azure Migrate does the heavy lifting for server estates. A lightweight appliance runs on premises, reads the inventory and sends metadata to an Azure Migrate project. Microsoft’s appliance documentation says one VMware appliance handles up to 10,000 servers across up to 10 vCenter Servers.
Hyper-V has its own ceiling of 5,000 servers and 300 hosts per appliance. The Hyper-V scale guide also allows up to 35,000 servers to be assessed in one project. So for the 400-VM estate in our example, a single appliance is plenty.
Assessments then size each server for Azure, flag readiness problems and feed the business case. Dependency analysis then maps which servers talk to each other, so tightly linked machines move in the same wave.
Microsoft recommends the agentless method. Agent-based analysis works only in the classic view as of October 2026 and takes no new servers. That view is due for deprecation by the end of 2026.
The agentless method installs nothing on the servers. According to Microsoft’s dependency analysis guide, the appliance polls TCP connection data every five minutes and sends it to Azure every six hours. Let it run across a full business cycle, such as a month-end close, before you fix the waves.
Assessment Settings That Shape the Business Case
Assessment settings deserve the same care, because they decide what the business case shows. Microsoft’s assessment properties page lists the options below, and finance should agree them before the first run.
| Setting | Options | What it changes |
|---|
| Sizing criteria | As-is on-premises or performance-based | Performance-based sizing reads CPU, memory, IOPS and throughput data |
| Performance history and percentile | A time window and a percentile of the samples | Applies only to performance-based sizing |
| Comfort factor | A multiplier on CPU, RAM, disk and network | A 10-core VM at 20% use sizes to two cores, or four at a factor of 2.0 |
| Savings option | None, reservations or a savings plan, for 1 or 3 years | None prices compute at pay-as-you-go rates for 730 hours a month |
| VM uptime | Defaults to 31 days a month and 24 hours a day | Lower it for servers that do not run around the clock |
| Azure Hybrid Benefit | On or off | On drops Windows and Linux license prices from VM costs for eligible licenses |
| Security | Defaults to Microsoft Defender for Cloud | Adds security readiness and cost to each VM |
Migration itself covers VMware, Hyper-V and physical servers. Microsoft’s overview adds “VMs from other public clouds, by treating them as physical servers.” That is the path teams take when migrating from AWS to Azure. Server migration is free for each machine’s first 180 days, then costs $25 per month for every instance still replicating.
Kanerika Service
Data and Cloud Migration Services
Kanerika plans and runs Azure migrations end to end, from discovery and business case to landing zone, cutover and data validation, with accelerators for the pipelines and reports native tools do not convert.
Explore Migration Services
Azure Database Migration Service and the End of Data Migration Assistant
Databases get a service of their own, and our primer on Azure data migration covers the planning around it. Azure Database Migration Service (DMS) moves SQL Server to three targets. Azure SQL Database is offline only, while Azure SQL Managed Instance and SQL Server on Azure VMs support online or offline moves.
Online mode keeps the source running while changes replicate, so downtime shrinks to the final cutover. Offline mode is simpler, but the database stays down for the whole copy. Pick online for systems that cannot pause, but offline for test databases or anything with a weekend window.
Mode also sets the price. Microsoft’s DMS pricing page lists a free Standard tier for offline moves, while online moves need the Premium tier, billed hourly per vCore. Each new Premium 4-vCore service is free for its first 183 days, or about six months.
Assessment has also moved. Microsoft retired Data Migration Assistant on July 16, 2025, so from that date SQL Server readiness and right-sizing come from Azure Migrate. MySQL and PostgreSQL discovery and assessment sit there as well, in public preview since February 2025 and September 2025.
DMS itself, according to its current overview (February 2026), covers SQL Server targets only. Older guides that still list Oracle, MySQL or PostgreSQL as DMS sources describe the older DMS (classic).
Target choice often moves the bill more than tool choice does. Compare service tiers with our guide to Azure SQL Database pricing before you send every database to Managed Instance by default. Our comparison of Azure SQL vs SQL Server explains when Database, Managed Instance or a VM fits each workload.
Azure Data Box, Storage Mover and AzCopy for Files and Bulk Data
File data is where teams most often pick the wrong tool, because all three options work and only one fits a given timeline. The deciding numbers are the volume and the bandwidth you can actually use, as well as the date the data must land.
Azure Data Box is the offline route. Microsoft ships a device, your team copies data onto it locally, and then the data is uploaded inside an Azure data center. The next-generation devices hold 120 TB and 525 TB of usable space. Data Box Heavy is retired, while the 80 TB device is retiring region by region.
Azure Storage Mover is the managed online route for shares. It moves SMB 2.x and 3.x shares to Azure Files or Blob storage and NFS shares to Azure Files. It also copies AWS S3 buckets to Blob. Sources on AWS FSx for SMB and Google Cloud Storage are in preview as of October 2026.
AzCopy is the command-line utility for copies to, from or between storage accounts. Microsoft points migrations above 1 TB to Storage Mover, so AzCopy suits scripted jobs and delta syncs, such as this typical v10 command:
# Copy a local share into a Blob container, including subfolders
azcopy copy "/mnt/finance-share" "https://<account>.blob.core.windows.net/finance?<SAS-token>" --recursive
Compared to each other, the three options look like this:
| Option | Mode | Best for | Watch-out |
|---|
| Azure Data Box | Offline, shipped device | Tens to hundreds of terabytes over a slow or shared link | Order fees and shipping time |
| Azure Storage Mover | Online, managed service | Share migrations when you want each run tracked as a job | Some sources are in preview (October 2026) |
| AzCopy | Online, command line | Scripted or one-off copies, and delta syncs after a Data Box load | Your team owns retries, logging and scheduling |
Delta Syncs After a Data Box Load
For the delta pass after a Data Box load, AzCopy’s sync command copies only new or changed files. A dry run lists those changes first, and a bandwidth cap protects the shared link:
# Preview the delta first, then sync it with a 500 Mbps cap
azcopy sync "/mnt/finance-share" "https://<account>.blob.core.windows.net/finance?<SAS-token>" --dry-run
azcopy sync "/mnt/finance-share" "https://<account>.blob.core.windows.net/finance?<SAS-token>" --delete-destination=false --cap-mbps 500 --log-level WARNING
Leave the delete-destination flag at false until validation passes, so a file removed at the source is not deleted in Azure.
Site Recovery, Resource Mover and Partner Tools: Where They Fit
Three more names show up on most tool lists, and each fits a narrower case than the lists suggest. Azure Site Recovery replicates machines for disaster recovery. Some teams use it to migrate, yet its pricing shows its purpose: each protected instance is free for 31 days, then billed per instance.
Azure Resource Mover moves Azure resources between Azure regions. It brings nothing in from on premises, so it belongs in a region consolidation project rather than a first migration.
Partner tools fill the gaps the native set leaves. The Azure Migrate overview (classic view) lists integrated tools from Carbonite, Cloudamize, CloudSphere, Corent, Device42, Lakeside, RackWare, Turbonomic, CloudRecon and Zerto. These can add license fees of their own, so the “no additional charge” line covers Microsoft’s tools only.
Talk to Kanerika
Mapping Your Estate to the Right Tools?
Kanerika reviews your servers, databases and file data, then shows which Azure tool fits each workload, what it will cost and where native tooling stops.
Schedule a Demo →
For web apps, Microsoft also provides the standalone App Service Migration Assistant, which moves .NET web apps from Windows servers to Azure App Service one site at a time. Azure Migrate can also assess the same apps first and point to App Service or Azure Kubernetes Service as the target. For ASP.NET apps on VMware, it then migrates them to App Service in bulk, without agents.
What Changed in Azure Migration Tools in 2025-2026
Microsoft’s migration tooling changed shape across 2025 and 2026, and many older guides still name retired tools. The dated list below comes from the Azure Migrate release notes and Microsoft’s product pages, checked in October 2026.
| Date | Change | What it means for your plan |
|---|
| March 1, 2024 | Movere retired | Use Azure Migrate discovery for inventory |
| February 2025 | MySQL discovery and assessment in Azure Migrate (public preview) | Assess MySQL alongside the servers it runs on |
| July 16, 2025 | Data Migration Assistant retired | Run SQL Server assessments in Azure Migrate, then move with DMS |
| September 2025 | PostgreSQL discovery and assessment (public preview) | Pilot preview features before you rely on them |
| March 2026 | Copilot migration agent (public preview) | Helps plan VMware migrations; execution stays in the Azure Migrate portal |
| June 2026 | GitHub Copilot modernization CLI integration (public preview) | Connects app code modernization to the migration project |
| September 2026 | Auto Wave Planning (public preview) | Drafts migration waves from inventory, assessments and dependencies |
| As of October 2026 | Data Box Heavy retired; 120 TB and 525 TB devices; 80 TB device retiring by region | Check device sizes and regional availability when you order |
| By the end of 2026 | Azure Migrate classic view scheduled for deprecation | Move agent-based dependency analysis to the agentless method |
Preview status still matters for planning. A preview feature can change before general availability. So treat the Copilot agent and Auto Wave Planning as planning aids, and confirm each wave by hand.
The retirements matter just as much, because they break old runbooks. A plan that still starts with Data Migration Assistant or Movere needs rewriting before the first appliance goes in.
How to Choose the Right Azure Migration Tool for Each Workload
Choosing well starts with three questions about each workload: what it is, how much data it carries and how long it can be down. The answers then point to a tool. They also point to one of the eight strategies in Microsoft’s Cloud Adoption Framework: retire, retain, rehost, replatform, refactor, rearchitect, rebuild or replace.
Kanerika’s podcast steps back from individual tools to the types of data migration every enterprise should know. It is a useful frame before the workload decisions below.
Two workloads show how far apart the answers can be. A file server with 5 TB of archives, a 1 Gbps link and a free weekend is a Storage Mover or AzCopy job. At 50 TB, the same link needs a Data Box. But a payments database that cannot stop needs DMS in online mode and a rehearsed cutover.
Decision Table: Workload, Downtime and Tool
The table below applies those three questions to the workloads that turn up in most estates.
| Workload | Downtime tolerance | Tool | CAF strategy |
|---|
| VMware or Hyper-V virtual machines | Hours, per wave | Azure Migrate server migration | Rehost |
| Physical servers, or VMs on AWS or Google Cloud | Hours, per wave | Azure Migrate, treated as physical servers | Rehost |
| Mission-critical SQL Server | Minimal | DMS online to SQL Managed Instance | Replatform |
| Dev and test SQL Server | Hours to a weekend | DMS offline | Replatform |
| .NET web apps | Hours | Azure Migrate web app migration in bulk; App Service Migration Assistant for one-off sites | Replatform |
| File shares that fit the network window | Online, with a final sync | Storage Mover or AzCopy | Rehost |
| Bulk data the link cannot carry in time | Days, during shipping | Azure Data Box | Rehost |
| Azure resources in the wrong region | Planned window | Azure Resource Mover | Not a CAF strategy (region move) |
| Apps nobody uses | Not applicable | None: decommission | Retire |
One habit saves weeks: work out the transfer time before you pick a data tool. Divide the data volume by the bandwidth you can really use rather than the rated link speed, then compare the result with the deadline. If the answer runs past the deadline, order a Data Box.
Worked Example: Moving a Mixed Estate
Back to the Friday plan. Here is how the four workloads map to tools, with the reasoning behind each choice.
- 400 VMware VMs: deploy one Azure Migrate appliance, well under its 10,000-server limit. Run agentless dependency analysis, build the business case, then migrate in waves with Azure Migrate server migration.
- 30 SQL Server instances: assess them in Azure Migrate for readiness and target size. Then move the critical ones with DMS online to SQL Managed Instance and the rest offline.
- 200 TB of file shares: at full line rate on 1 Gbps, the copy takes about 18.5 days, and at a realistic 70% use it takes about 26 days. Load the bulk onto one 525 TB Data Box or two 120 TB devices, then sync changes with Storage Mover or AzCopy.
- .NET web apps: assess them in Azure Migrate, then migrate them to App Service in bulk with Azure Migrate’s agentless web app migration. Keep App Service Migration Assistant for one-off sites.
The math behind item 3 is simple, using decimal units. 200 TB is 1.6 million gigabits, so the copy needs 1.6 million seconds at 1 Gbps, or about 18.5 days. Shared links rarely hold full rate during business hours, so 70% is the safer planning figure.
Data Box wins here for a second reason as well. The same link must carry server replication for 400 VMs, so a month of file copying would slow every wave behind it.
How an Azure Migration Runs, Step by Step
An Azure migration runs as a repeatable sequence, and each step has a native tool behind it. Teams that skip a step usually pay for it later, at cutover. A written data migration framework keeps that sequence the same from one wave to the next.
- Discover: deploy the Azure Migrate appliance and build a full inventory, dependencies included.
- Assess and build the business case: size targets, flag blockers and compare on-premises and Azure costs.
- Prepare the landing zone and guardrails: set up subscriptions, networking, identity, policy and budgets before the first workload lands.
- Pilot: move one low-risk application end to end before any wave, and time every step.
- Plan and run waves: group servers by dependency, and let Auto Wave Planning (preview) draft the groups if you use it.
- Cut over with rollback criteria: agree in advance what failure looks like and how to switch back.
- Validate and optimize: test data and applications, right-size, then decommission the source.
The pilot is the step teams most often cut, even though it is the cheapest insurance in the plan. One application moved end to end shows real replication times, network gaps and sign-off delays before 400 servers depend on them. A phased cloud migration roadmap puts that pilot on the calendar ahead of any wave date.
Landing Zone Guardrails Before the First Wave
Landing zone work is rushed far too often. Microsoft’s Cloud Adoption Framework defines an Azure landing zone as the governed, multi-subscription foundation that workloads land in. Policy, identity and budget alerts are far easier to set there before workloads arrive than to retrofit across hundreds of machines.
A region guardrail is a sensible first policy. This Azure CLI command assigns Microsoft’s built-in Allowed locations policy, whose default effect is Deny, so nothing deploys outside the regions you list:
# Built-in policy "Allowed locations" (default effect: Deny)
az policy assignment create \
--name "migration-allowed-regions" \
--policy "e56962a6-4747-49cd-b67b-bf8b01975c4c" \
--scope "/subscriptions/<subscription-id>" \
--params '{"listOfAllowedLocations": {"value": ["eastus2", "centralus"]}}'
Keep guardrails like this in code rather than in portal clicks. Our Azure Terraform guide shows how teams manage that infrastructure as code.
Validating Data and Cutover
Cutover is where tooling ends and judgment begins. Azure Migrate replicates servers and DMS moves databases, but neither proves that the business data arrived intact.
Plan row counts, checksums and report-level comparisons before the first wave, then run them on every wave. Our guide to data migration testing covers the test types in detail. The companion piece on risks in data migration lists the failure points worth planning against.
Write the rollback criteria down before the first wave. A sound rule keeps the old system ready until validation passes and the business owner signs off. Our data migration checklist turns these checks into sign-off items for each wave.
Cutover timing deserves the same care. Pick a window when transaction volume is low and freeze changes on the source. Then keep the people who can approve a rollback on the call until the first business day closes cleanly.
What Azure Migration Tools Cost
Tool charges are the smallest line in an Azure migration budget, which surprises many finance teams. Microsoft charges nothing for Azure Migrate discovery and assessment, while the paid parts start only after free windows.
| Tool | Free window | After the window | Watch-out |
|---|
| Azure Migrate discovery and assessment | Always | No additional charge | Partner tools inside it may charge |
| Azure Migrate server migration | First 180 days per machine | $25 per month per replicated instance | Slow waves can cross the 180-day mark |
| Azure Site Recovery | First 31 days per protected instance | Per-instance charge (see the pricing page) | Disaster recovery use is ongoing, not one-off |
| Azure Database Migration Service | Standard tier always; Premium 4-vCore for 183 days | Premium billed hourly per vCore | Online moves need Premium |
| Azure Storage Mover | All current features | Target storage capacity and transactions | Transactions scale with file count, not size |
| Agent-based dependency analysis | Classic view only | No new servers; Log Analytics charges on existing ones | Use agentless analysis |
| Azure Data Box | None | Order and shipping fees | Plan for device lead time |
The Costs Outside the Tool Line
The real costs sit outside the tool line. Dual running is usually the largest, because the old data center and the new Azure footprint both bill until cutover ends. How much of the estate stays hybrid decides how long that dual running lasts, so settle the scope early. Egress fees apply when data leaves another cloud, and storage transactions add up during large copies.
Microsoft’s Storage Mover billing guide says transaction counts scale with the number of files and folders, not their size. So 1 GiB of small files needs more transactions than 1 GiB of large ones. Each repeat copy run may also need as many transactions as the first.
Waste is the other risk. Flexera’s 2026 State of the Cloud report puts wasted spend at 29% of IaaS and PaaS budgets, up after five years of decline. The survey drew 753 respondents. The same survey also found that 73% run a hybrid cloud.
Right-sizing during assessment is the cheapest fix, and our guide to Azure cost optimization covers the levers that follow. Large programs should also review a Microsoft Azure Consumption Commitment (MACC) before the first wave, since migration spend is large and predictable.
Limits of Native Azure Migration Tools and Common Mistakes
Native tools are strong at moving servers and databases, but they stop at clear boundaries. Knowing those boundaries early saves a re-plan halfway through the program.
Scale is the first boundary: the appliance and project limits covered above mean very large estates need several appliances, and sometimes several projects.
Execution is the second. The Copilot migration agent helps plan VMware migrations, yet replication and cutover still run in the Azure Migrate portal. Code is the third: Azure Migrate does not refactor applications or convert SSIS packages, SSAS models, or Crystal and Cognos reports.
Some workloads should not be lifted and shifted at all. Hardware-bound systems, latency-sensitive plant systems and apps due for retirement may belong in a hybrid or data center modernization plan instead.
Licensing is a quieter boundary. Moving a server does not move the right to run the software on it. Check vendor license terms and support status for each major application before its wave is scheduled. For Microsoft software, our guide to Microsoft licensing optimization covers that review in more depth.
Five Mistakes That Cost Weeks
The same five mistakes come up again and again:
- Skipping dependency analysis, then splitting tightly linked servers across waves.
- Treating Site Recovery as the default migration tool when Azure Migrate fits better.
- Copying 200 TB over the network before working out the transfer time.
- Leaving machines replicating until well past the 180-day free window.
- Assuming the tools validate business data, when they only move it.
Each of these costs weeks, not hours, and each can be caught during assessment. The fix is the same every time: turn the risk into a named check with an owner. Our wider list of cloud migration best practices covers the habits that prevent them.
Dependency maps get reviewed before waves are fixed. Transfer times are worked out before a data tool is chosen. Someone also reads the replication report every month, so no machine drifts past its free window.
A quick estimate of savings and effort before the first wave also makes the business case easier to defend.
Migration ROI Calculator
Estimate Your Azure Migration ROI
Put rough numbers on effort, savings and payback before the first wave with Kanerika’s Migration ROI Calculator.
Calculate Migration ROI →
How Kanerika Runs Azure Migrations
Kanerika plans and runs Azure migrations end to end, using native Microsoft tools where they fit and its own accelerators where they stop. In December 2025, Microsoft awarded Kanerika the Data Warehouse Migration to Microsoft Azure specialization.
Azure Migrate moves servers well, but it does not convert the data estate running on them. SSIS packages, SSAS models and Crystal or Cognos reports still have to be rebuilt on the target platform.
That is where Kanerika’s FLIP accelerators sit. FLIP automates the conversion of pipelines, models and reports, and it also runs validation alongside. Teams then check converted output against the source instead of rebuilding by hand.
Each engagement starts the same way. Azure Migrate discovery and dependency analysis come first, so every workload has a named tool, a wave and a validation test before anything moves. Data checks then run inside each wave rather than at the end of the program.
Teams heading further, to a lakehouse, can follow our Azure to Fabric migration guide. Kanerika’s data and cloud migration services cover the full path, from assessment and landing zone to cutover and validation. After go-live, Azure managed services keep patching, monitoring and cost reviews on a schedule.
One client shows the pattern. A 96-year-old North American steel mill services firm ran a costly legacy system with weak encryption. Kanerika then moved it to Dynamics AX, Azure Data Lake and Power BI. The case study reports 25% lower storage expenses, a 40% gain in productivity and a 2.5% rise in revenue.
Case Study
25% Lower Storage Costs for Steel Manufacturing on Azure
A 96-year-old North American steel mill services firm replaced a costly legacy system with Dynamics AX, Azure Data Lake and Power BI, cutting storage expenses by 25% and improving productivity by 40%.
Read the Case Study →
Which Azure Migration Tool Should You Start With?
Start with Azure Migrate, whatever the estate looks like. Discovery, dependency analysis and the business case cost nothing, and together they tell you which specialist tools you need.
From there, the routing is short:
- Servers, including VMs on AWS or Google Cloud: Azure Migrate server migration.
- SQL Server: Database Migration Service, online for systems that cannot stop.
- Files and bulk data: Storage Mover or AzCopy online, and Data Box when the link is too slow.
- .NET web apps: Azure Migrate in bulk, or App Service Migration Assistant for one-off sites.
- Disaster recovery and later region moves: Site Recovery and Resource Mover.
So which Azure migration tool moves which workload? Azure Migrate’s free assessment decides the order first. Then Azure Migrate moves the servers, DMS moves the databases, and Data Box or Storage Mover moves the files.
Frequently Asked Questions
What are Azure migration tools?
Azure migration tools are the Microsoft services that discover, assess and move workloads into Azure. Azure Migrate is the hub for servers, web apps and assessments. Azure Database Migration Service moves SQL Server databases, while Data Box, Storage Mover and AzCopy move files and bulk data. Site Recovery and Resource Mover cover disaster recovery and moves between Azure regions.
What is Azure Migrate used for?
Azure Migrate is used to discover on-premises and other-cloud workloads, assess their readiness for Azure and migrate servers. It builds an inventory with a lightweight appliance, maps dependencies, right-sizes targets and produces a business case comparing on-premises and Azure costs. It migrates VMware, Hyper-V and physical servers, and it assesses SQL Server instances and web apps.
What are the steps of an Azure migration?
A typical Azure migration runs in seven steps. Discover the estate with Azure Migrate, assess it and build a business case, then prepare a landing zone with guardrails. Pilot one low-risk application, plan and run migration waves, and cut over with written rollback criteria. Finally, validate data and applications, right-size resources and decommission the source.
How long does an Azure migration take?
Duration depends on estate size, dependencies, data volume, downtime tolerance and how many waves you run. Transfer time is often the hidden driver. For example, 200 TB over a 1 Gbps link takes about 26 days at 70% utilization, which is why Data Box often wins. Running a pilot gives you real timings to plan the remaining waves.
How much does an Azure migration cost?
Tooling is the smallest cost. Azure Migrate assessment is free, and server migration is free for 180 days per machine, then $25 per month per replicated instance. The larger costs are dual running until cutover, egress from other clouds, Data Box order fees and idle capacity. Flexera’s 2026 report puts wasted cloud spend at 29%.
What are the benefits of migrating to Azure?
Moving to Azure replaces capital spending on hardware with operating spending you can scale up or down. It gives access to managed databases, analytics and AI services, and it can retire aging data centers. The benefits depend on right-sizing and governance, because unmanaged cloud estates waste money, as Flexera’s 29% waste figure shows.
Does Azure Migrate support dependency mapping?
Yes. Azure Migrate maps which servers communicate with each other, so tightly linked machines move in the same wave. Microsoft recommends the agentless method. Agent-based analysis now works only in the classic view and cannot onboard new servers. Existing setups also pay Log Analytics charges. Microsoft has scheduled that classic view for deprecation by the end of 2026.
Is Data Migration Assistant still supported?
No. Microsoft retired Data Migration Assistant on July 16, 2025. SQL Server readiness and right-sizing assessments now run in Azure Migrate, and the database moves themselves run in Azure Database Migration Service. Any runbook that still starts with Data Migration Assistant should be updated before your first migration wave begins.
Is Azure Migrate free?
Azure Migrate discovery and assessment are available at no additional charge. Server migration is free for the first 180 days per machine, then costs $25 per month for each instance still replicating. You still pay for the Azure resources you run, and third-party partner tools used inside Azure Migrate may charge their own fees.
What is the best Azure migration tool?
No single tool is best for every workload. Start with Azure Migrate, because its discovery, dependency analysis and business case are free and show what you need. Then use Azure Migrate for servers, Database Migration Service for SQL Server, and Storage Mover, AzCopy or Data Box for files, depending on data volume and bandwidth.
What is the difference between Azure Migrate and Azure Site Recovery?
Azure Migrate is built for migration, covering discovery, assessment, business cases and server moves. Azure Site Recovery is built for disaster recovery: it replicates machines so they can fail over if a site goes down. Site Recovery can move machines, but its per-instance billing after 31 days reflects ongoing protection, so Azure Migrate is usually the better fit.
What is the difference between Azure Resource Mover and Azure Migrate?
Azure Resource Mover moves resources that already run in Azure from one Azure region to another. Azure Migrate brings workloads into Azure from on-premises data centers or other clouds. Use Azure Migrate for the initial migration, and use Resource Mover later if you need to consolidate or relocate resources between Azure regions.
Which tool migrates SQL Server databases to Azure?
Azure Database Migration Service moves SQL Server to Azure SQL Database offline, to Azure SQL Managed Instance online or offline, and to SQL Server on Azure VMs online or offline. Assessment happens in Azure Migrate, which checks readiness and sizes targets. Data Migration Assistant used to handle assessments, but Microsoft retired it on July 16, 2025.
Can Azure Migrate move VMware virtual machines?
Yes. Azure Migrate discovers, assesses and migrates VMware virtual machines. One VMware appliance can discover up to 10,000 servers across up to 10 vCenter Servers, so most estates need only one. Teams typically run agentless dependency analysis first, group linked machines into waves, then replicate and cut over each wave from the Azure Migrate portal.
Can Azure Migrate move workloads from AWS or Google Cloud?
Yes. Azure Migrate migrates virtual machines from other public clouds by treating them as physical servers. You discover and assess them like any physical machine, then replicate them to Azure. For object storage, Azure Storage Mover can copy AWS S3 buckets to Blob storage, and Google Cloud Storage sources are in preview.
What are the 8 Rs of cloud migration?
Microsoft’s Cloud Adoption Framework lists eight migration strategies: retire, retain, rehost, replatform, refactor, rearchitect, rebuild and replace. Rehosting moves a workload as it is, replatforming makes small changes such as moving to a managed database, and the later strategies change more of the application. Older guides describe six or seven Rs, so check which list you follow.
