An in-place upgrade to Windows Server 2025 keeps an existing server installation while replacing its operating system. It can reduce migration work, but it carries existing configuration, application dependencies and some historical problems forward. Plan it as a recoverable service change with a compatibility decision, tested backup and explicit acceptance checks, rather than as an installer that is expected to preserve everything automatically.
This article helps you decide whether an ordinary nonclustered member server is a suitable candidate and prepare the upgrade. Domain controllers, clustered workloads and specialized roles need their own migration or rolling-upgrade procedures. Microsoft's current guidance recommends introducing new domain controllers on a clean OS and retiring older controllers rather than using in-place upgrade for AD DS.
Choose upgrade or migration based on the workload
List every role, application, agent, scheduled task, certificate and external dependency on the server. Include backup software, monitoring agents, endpoint protection, licensing services and vendor utilities. An application that runs on the old Windows release is not automatically supported on Windows Server 2025.
Ask each application owner for a support statement and an acceptance test. If an unsupported component cannot be replaced, the OS upgrade is not ready. A side-by-side migration may provide a cleaner rollback because the original server remains available, though data synchronization and cutover still need careful design.
Prefer migration when the current installation has unexplained corruption, undocumented security settings or obsolete applications nobody can validate. In-place upgrade does not remove the need to understand those problems. A server carrying a critical single-instance application deserves a stronger recovery plan than an easily rebuilt management utility.
Verify the exact supported path
Microsoft documents direct media-based upgrades to Windows Server 2025 from Windows Server 2012 R2 and later for qualifying nonclustered systems. Check the current table for the exact source, destination, edition, licensing channel and method. Windows Server 2012 without R2 is not the same source as Windows Server 2012 R2.
Restrictions include language changes and switching between Server Core and Desktop Experience during the upgrade. Edition changes have their own rules; an upgrade is not a way to downgrade Datacenter to Standard. Evaluation media, prerelease builds, VHD-boot installations and specialized editions require separate scrutiny. The option to keep applications and settings must be available for the intended supported path.
Windows Update and installation-media methods have different prerequisites. The Windows Update path for Windows Server 2019 and 2022 requires specified cumulative updates and other conditions, which can change through servicing. Verify the current prerequisite list immediately before scheduling rather than preserving an old knowledge-base number as a permanent requirement.
Confirm platform and licensing readiness
Check the Windows Server 2025 hardware requirements and the hardware or virtualization platform's support. Validate firmware, storage-controller drivers, network drivers and virtual integration components. Minimum installable resources are not necessarily enough for the production workload after upgrade.
For hosted virtual machines, confirm the provider's supported upgrade method, console access and recovery process. A generic media procedure may not apply to a particular cloud platform. Ensure the administrator can reach a recovery console if networking or RDP fails. Test access before the maintenance window, not after the remote session disappears.
Confirm the target edition, license entitlement, activation method and any application licensing effects. Retain trusted installation media with the correct language and edition. Do not start Setup with evaluation media just because it is convenient to obtain when the required production upgrade path calls for licensed media.
Collect a useful baseline
Save diagnostic and configuration information to a protected location off the server. The following examples collect an inventory; adapt the destination to a real controlled share:
$dest = '\\backup01\UpgradeRecords\APP01'
New-Item -ItemType Directory -Path $dest -Force
Get-ComputerInfo | Out-File "$dest\ComputerInfo.txt"
Get-WindowsFeature | Where-Object Installed |
Export-Csv "$dest\InstalledFeatures.csv" -NoTypeInformation
Get-NetIPConfiguration | Format-List * |
Out-File "$dest\Network.txt"
Get-Service | Select-Object Name,Status,StartType |
Export-Csv "$dest\Services.csv" -NoTypeInformation
systeminfo | Out-File "$dest\SystemInfo.txt"
Also record application versions, bindings, certificates and private-key recovery arrangements, firewall exceptions, storage mappings, service identities, scheduled tasks and current monitoring health. Protect this material because it describes the infrastructure. An inventory should help rebuild or diagnose the server, not expose secrets in a shared folder.
Review existing System and Application errors. Resolve relevant disk, filesystem, component servicing and application failures before upgrading. Otherwise it becomes difficult to tell whether an error is new or merely noticed after the OS change.
Prove the backup can restore the service
Take a full backup covering the operating system, applications and data, with the workload's supported consistency method. For databases, identity services and replicated applications, follow the application's recovery requirements in addition to image-level protection. Store the backup away from the server and confirm retention exceeds the change window.
Perform a restore test in a suitable isolated environment and record the elapsed time. A successful backup job is evidence that data was captured, not that the application can be recovered within its required outage. Ensure recovery credentials, encryption keys and installation dependencies are available to the people conducting the change.
Do not promise that an automatic OS uninstall option will provide the rollback. Recovery after a server upgrade may require restoring the complete system and reconciling data written after the backup. Set a decision deadline for rollback based on that recovery time and the application's data-loss tolerance.
Prepare the maintenance window
Apply approved source-OS updates and finish pending restarts. Check free space on system and data volumes, stop or quiesce the application as documented and suspend conflicting jobs. Coordinate endpoint protection and backup-agent actions with their supported upgrade instructions rather than disabling safeguards indiscriminately.
If NIC Teaming is configured, review Microsoft's requirement to disable it before an in-place upgrade and plan how the machine remains reachable. Record the team and adapter configuration and arrange console access. A remote-only change that removes its own network path is not ready to execute.
Notify dependent service owners of the actual outage and validation period. Reserve time for application testing and a possible restore, not just the optimistic Setup duration. Confirm who can approve acceptance, extend the window or invoke recovery.
Run Setup only after the readiness decision
Start the supported upgrade method from the running source OS and select the correct target edition and installation option. Review compatibility findings and the option to retain applications and settings. If the retention choice is unavailable, stop and investigate the media, source state and supported path instead of proceeding with a destructive alternative.
Follow Setup prompts and allow its required restarts. Keep console access available and avoid interrupting a progressing upgrade because RDP is temporarily unavailable. If Setup fails, preserve its logs and exact error before rerunning or restoring. Repeated blind attempts can overwrite useful evidence and consume the recovery window.
Accept the service through defined checks
After upgrade, confirm the OS build, activation, device health, network settings and installed roles. Apply appropriate approved updates and complete restarts. Compare service states and scheduled tasks with the baseline. Check backup, monitoring and endpoint protection agents as well as the main application.
Run representative business transactions from a real client, including authentication and external dependencies. For a file service, inspect access and permissions; for an application service, verify reads, writes and background jobs. A working desktop or RDP session is not application acceptance.
Keep enhanced monitoring through the agreed observation period. Retain the pre-upgrade backup until the owners accept the result and a new post-upgrade backup has been verified. If acceptance fails, invoke the documented recovery process before the rollback deadline, accounting for any new data created since the original backup.
Consult Microsoft's upgrade-path overview and current in-place upgrade procedure immediately before implementation.
Related procedures
How to Migrate a Windows File Server with Storage Migration Service; How to Replace an Old Domain Controller with Windows Server 2025; How to Migrate IIS Websites to a New Windows Server.