Hyper-V Virtual Machine Management Service Stuck: Fix VMMS Not Starting or Stopping

A Hyper-V virtual machine management service stuck in Stopping, or a host that sits on “Shutting down service: Hyper-V Virtual Machine Management”, is a host-side symptom, even when the root cause turns out to be one VM. The Virtual Machine Management Service (vmms) is the host service that manages VMs, takes checkpoints, runs migrations and Replica, and answers Hyper-V Manager. When it hangs or won’t start, management stops. Running VMs normally keep running: Microsoft documents that VMs stay up with no management access while VMMS is down. So the first job is to confirm scope before changing anything. Start with read-only checks of the service state and the VMMS and System event logs, then decide whether the cause is the service itself, one VM’s worker process, storage, a pending operation, or remote access.

Quick Answer: Check VMMS Without Disrupting Running VMs

When the Hyper-V virtual machine management service is stuck, every command in this section is safe to run first: all of them are read-only. Run them on the Hyper-V host in an elevated PowerShell session.

sc.exe queryex shows the state the Service Control Manager (SCM) sees, plus the process ID. A Running state is not proof that VMMS is healthy. A service can report Running while it is blocked and ignores management calls.

One distinction matters before anything else. A host shutdown screen is not a guest shutdown screen. If the “Shutting down service” text is inside a VM console window, the problem is in the guest OS, not in VMMS.

What Is the Hyper-V Virtual Machine Management Service (VMMS)?

VMMS is the host service, display name Hyper-V Virtual Machine Management, that orchestrates Hyper-V on that host. Its process is vmms.exe. It handles VM creation and deletion, start and stop requests, checkpoints, live migration, Replica, and the WMI and PowerShell interfaces Hyper-V Manager uses. It also hosts the Microsoft Hyper-V VSS Writer that host-level Hyper-V backups depend on.

Several similarly named components are easy to mix up:

What happens when VMMS stops is documented. Microsoft’s Best Practices Analyzer guidance, The Hyper-V Virtual Machine Management service must be running, says running VMs continue to run, but you can’t manage, create, or delete VMs until the service is back. The VMMS event log records the same thing when the service stops with VMs running: “Hyper-V Virtual Machine Management service is shutting down while some virtual machines are running. All running virtual machines will remain running with no management access.”

That’s why a VMMS problem is usually a management outage first and a workload outage second. In-flight operations are a different matter. A backup, checkpoint merge, migration, or Replica cycle that VMMS was coordinating may fail, abort, or need follow-up verification.

Hyper-V Stuck at “Shutting Down Service: Hyper-V Virtual Machine Management”

This is the symptom most people search for, and it isn’t always a deadlock. During host shutdown, VMMS carries out each running VM’s Automatic Stop Action before it stops. When a VM is configured with the Save action, VMMS writes that VM’s memory to disk. A host with many running VMs, large amounts of assigned memory, or slow storage can sit on that screen for a long time while doing exactly what it was configured to do. The available stop actions are documented in the Set-VM reference.

The three stop actions behave differently:

• Save: VMMS saves each VM’s memory to disk. Time scales with the memory being saved and storage speed.
• Shut down the guest operating system: VMMS asks each guest to shut down through integration services. A guest that ignores the request, or is busy with its own updates, holds up the host.
• Turn off: fastest, and the equivalent of pulling power on each VM. Not a safe choice for workloads that hold data in memory.

Slow stop actions are only one possibility. The same status line can also sit on screen because of a VM whose worker process hung, a backup or Replica operation still running, a live migration in progress, storage that has gone slow or unreachable, or Windows servicing that makes the whole shutdown slow. A long update shutdown is not automatically a VMMS hang.

Check stop actions before a planned shutdown, while management still responds:

Once the host is already on the shutdown screen, you usually can’t open a new session to check anything. Record when the shutdown started and which VMs you know were running, check whether a backup window or Replica cycle overlaps, and decide whether the workloads can take an unplanned outage.

A hard power-off at that screen has real consequences. Every VM on the host loses power at once. Saved-state files being written can be left incomplete. On a cluster, the node drops without draining and the cluster has to fail over its roles. Treat it as an incident decision, not a troubleshooting step.

What to check after the host comes back

If the shutdown screen prevented collecting evidence, gather it after the restart. These logs reconstruct what happened. A current Get-VM result shows the state now, not the state during the hung shutdown.

Look for the last VM VMMS was working on before the hang. Errors about saving a VM, or a worker that stopped responding, point at that VM. Before the next planned shutdown, stop or save that VM yourself and watch whether it completes. If slow saves are the pattern, review each VM’s Automatic Stop Action. Changing it is a configuration decision, covered in Hyper-V VM configuration.

Hyper-V Virtual Machine Management Service Stuck Stopping

STOP_PENDING means the SCM asked VMMS to stop and VMMS hasn’t finished. That’s different from a service that is stopped, and from one that reports Running but doesn’t respond.

When the Hyper-V virtual machine management service is stuck stopping, work out what it is waiting for:

If one VM is clearly the blocker, the per-VM recovery path, including how to match that VM to its own vmwp.exe, is in Hyper-V VM won’t start. Terminating the right worker process is a hard power-off for that one VM. Terminating the wrong one is a hard power-off for a VM that was fine.

Terminating vmms.exe is not a routine fix, and it isn’t equivalent to a controlled stop. Running VMs normally survive it, but anything VMMS was in the middle of is abandoned without cleanup, and the underlying cause, such as the hung worker or the storage wait, is still there when the service comes back. If you do terminate it, use the exact PID from sc.exe queryex, not an image-name kill, then go straight to the verification checklist below.

VMMS Failed to Start, Not Running, or Missing

Start by confirming the service exists and how it’s configured:

If that returns nothing, go to the missing-service branch below. If the service exists, read its dependencies on this host and the SCM events:

When the Hyper-V Virtual Machine Management service failed to start, read the result by branch:

• Startup type is Disabled: find out what changed it first, such as a hardening baseline or a GPO, or it will change again. A Manual startup type isn’t automatically broken either; check whether someone set it deliberately. Once a misconfiguration is established, Microsoft’s Best Practices Analyzer guidance restores it as shown below.
• A dependency failed to start: use the dependencies -RequiredServices reports on this host, not a list copied from another server, and read that service’s own SCM error.
• The service starts and then stops: read the Hyper-V-VMMS-Admin log for the start attempt. A crash or configuration error is usually logged there.
• Hypervisor not running: VMMS can start while VMs still can’t run, if the hypervisor didn’t load at boot. That branch is covered with HypervisorPresent and hypervisorlaunchtype in the VM won’t start guide linked above.

Hyper-V Virtual Machine Management service missing

If the service isn’t found, it isn’t installed on the machine you’re querying. That commonly means one of two things: the Hyper-V role isn’t installed there, or you’re on a management workstation that has only the Hyper-V tools. Confirm you’re on the intended host first.

On Windows Server, check the role state:

On Windows client, Get-WindowsFeature isn’t available; check the optional feature instead with Get-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V. If the machine is supposed to be a Hyper-V host and the role is missing or partially installed, look at Windows servicing errors before any repair. Installing or repairing the role is a maintenance-window task that can require a restart, not a first step.

Permission errors

Search queries mention a “service does not have permission” variant, but there’s no single verified Microsoft error message behind it, and not every permission error is a VMMS service-account problem. Capture the full text, the Event ID, and the path or object that was denied. Then work out which of three cases it is: the service’s own access, access to VM storage, or authorization for a remote Hyper-V Manager user. Fix that specific principal on that specific resource.

Don’t change the VMMS service logon account, and don’t reset ACLs across the VM storage tree as a blanket fix. For VM storage, Microsoft’s Hyper-V installation, configuration, and operational failure guide focuses on access for Administrators, Hyper-V Administrators, and NT VIRTUAL MACHINE\Virtual Machines on VM folders and disks. A per-VM access-denied error (0x80070005) has its own section in the VM won’t start guide.

The same Microsoft guide includes WMI repository repair steps, but only for a diagnosed corrupt repository. A VMMS startup failure alone is not evidence of WMI corruption, and resetting the repository affects everything else on the host that uses WMI.

Hyper-V Manager Stuck Connecting to Virtual Machine Management Service

When Hyper-V Manager sits on “Connecting to Virtual Machine Management service”, the cause is either VMMS on the target host or the path between your console and the host. Prove which one before fixing either:

1. On the host itself, run Get-VM in an elevated session. If it returns quickly, VMMS is answering locally.
2. Open Hyper-V Manager locally on the host, if the host has a GUI. If local works and remote doesn’t, the problem is remote access.
3. Try from a second admin machine. One broken workstation points to that workstation’s configuration or credentials.

A quick Get-VM is good evidence, not proof under all conditions. A partly blocked VMMS can answer simple queries and hang on others. If Get-VM hangs on the host itself, treat it as a local VMMS problem and go back to the stuck-stopping and event log sections.

If VMMS is fine locally, the cause is remote management: WinRM, DCOM, firewall rules, delegation, or credentials. That’s a different troubleshooting path, and restarting VMMS won’t fix it. A related variant, the Move Wizard reporting that it can’t contact the Virtual Machine Management Service, belongs to migration troubleshooting in Hyper-V live migration failed.

How to Restart VMMS Safely and What It Affects

Recovery for a Hyper-V virtual machine management service stuck in any state sits on a risk ladder. Use the lowest step that resolves the problem:

What a VMMS restart can affect

• Running VMs: normally keep running, per Microsoft’s documentation and the VMMS shutdown event text. Verify it afterwards rather than assuming it.
• Live migrations and storage moves: an operation in progress may fail or abort. Check where the VM and its files ended up.
• Checkpoints and merges: a checkpoint being created or merged may need follow-up. Check the checkpoint and disk state afterwards.
• Backups: the Hyper-V VSS Writer runs inside vmms.exe, so a host-level backup in progress is likely to fail.
• Replica: a replication cycle in progress may be interrupted. Check replication health afterwards, as described in Hyper-V Replication.
• Clusters: check the node’s roles and cluster resource state before and after. A restart during a cluster operation adds a second problem to the first. Node draining and role behavior are covered in Hyper-V failover clustering.

When the impact check is clean and you’re in a change window:

Restart-Service can itself hang if VMMS stays in STOP_PENDING. That’s the signal to go back to what VMMS is waiting on, not to jump straight to a forced kill.

Review VMMS/Admin and Service Control Manager Events

The logs explain why the Hyper-V virtual machine management service is stuck. Collect both sources read-only:

Correlate by timestamp. Line up the SCM stop or start request, the last VMMS events before the hang, and any VM or storage errors in the same minutes. Read the message text, not only the ID. There’s no single “VMMS stuck” Event ID, and the same symptom can come from very different causes.

Two IDs come up often and get misread:

• Event 14050 (Hyper-V-VMMS) is a Service Principal Name registration failure. Microsoft lists it as its own symptom. It affects authentication for migration and remote management, but it isn’t a cause of VMMS stuck stopping unless the logged condition actually matches.
• Error ID 1602 comes from System Center Virtual Machine Manager. Its message, “The Virtual Machine Manager service on that server did not respond”, refers to the SCVMM service (vmmservice.exe), not Hyper-V VMMS. Microsoft documents it in The Virtual Machine Manager console doesn’t start and you receive error ID 1602. Troubleshoot SCVMM separately.

Troubleshooting Decision Table

The order below reflects practical investigation priority, not frequency.

Escalate to Microsoft Support when VMMS repeatedly fails to start with logged errors you can’t map to configuration, or hangs again after the blocker is cleared. Bring the VMMS-Admin and System logs, the sc.exe queryex output, and the VM states.

Frequently Asked Questions

Is it safe to restart the Hyper-V Virtual Machine Management Service?

Usually, inside a change window and after checking that no backup, migration, checkpoint, or Replica operation is running. Running VMs normally keep running. Management, and anything VMMS was in the middle of, is interrupted.

Does restarting VMMS shut down running virtual machines?

No, not normally. Microsoft documents that running VMs continue to run while VMMS is stopped, with no management access. Verify the VM states afterwards anyway.

Why is the Hyper-V Virtual Machine Management Service stuck stopping?

It’s waiting on something it has to finish first: a VM’s stop action, a hung worker process, a storage wait, or an in-flight operation such as a backup or migration. Identify the blocker with sc.exe queryex vmms, the VM states, and the VMMS-Admin log before forcing the stop.

How do I fix “Shutting down service: Hyper-V Virtual Machine Management”?

If VMs are configured to Save or shut down, VMMS may simply be working through them, and waiting is often the right move. For the next shutdown, stop or save the VMs first and review their stop actions. A hard power-off at that screen affects every VM on the host.

Why is Hyper-V Manager stuck connecting to the Virtual Machine Management Service?

Either VMMS on the host is unresponsive or the remote management path is broken. If Get-VM works on the host itself, it’s a remote-access problem, and restarting VMMS won’t help.

What if the Hyper-V Virtual Machine Management Service is missing?

It isn’t installed on the machine you’re querying: either the Hyper-V role isn’t installed there, or it’s a workstation with only the management tools. Confirm you’re on the intended host and check the role state before repairing anything.

What is the difference between vmms.exe and vmwp.exe?

vmms.exe is the single host management service. vmwp.exe is a worker process associated with an active VM. Stopping VMMS doesn’t stop VMs. Terminating a vmwp.exe hard-stops the VM it belongs to.

Final Thoughts

A Hyper-V virtual machine management service stuck in Stopping is far more often waiting on something than broken itself. Read the service state, the VM list while it still responds, and the logs first. Most of the time that points at a VM’s stop action, a hung worker, storage, or a backup still running. Restarting or killing the service comes after that, inside a change window, with in-flight operations accounted for.