how to fix broken VM with no remote access after July update,

@Mike Collins Did you tried by resizing the VM?

Can you tell me what 'compatible SKUs' are correct ? we opted for a default size of our VM's they are D4as_v6 size

I have just resized it to D4alds and still have the same issue

I have switched back to D4as_v6 and still the same

This July KB5062553 issue has been ongoing since last Friday now and we still cannot access our VM's is there any update to fix this properly.

Again, these have always been our security settings

Security

Security type Trusted launch

Enable secure boot Enabled

Enable vTPM Enabled

We have already applied the KB5064489 fix on 1 of our VM's

We have also tried resizing the same VM

I even have forced TCP only over the RDP

but every time we try and access the VM's by RDP the VM Agent stops and becomes in a Not Ready state, and after it enters the Not ready state we can no longer access the VM by serial console either. The Only way is to Stop and Start the VM and then the Agent becomes Ready and we can access it by Serial Console again.

This all only started happening after the July KB5062553 update was applied and is affecting both azure domain controllers, the azure exchange server and 2 other servers.

Hi Can anyone else help, so far I am getting almost nowhere. altho I have figured a little more out but it does not resolve the issue.

The Original VM had 3 disks attached.

I have taken a snapshot of the original OS Disk and created a new drive from it

I then built a new VM and swapped the OS drive with the one I built from the snapshot of the original VM.

This new VM boots fine and I can login at this point.

I then created snapshots of the 2 additional disks and created drives from them too.

If I attach them to the running new VM I can see all the data fine on them.

However, if I shutdown the new VM from the OS and then stop the VM in the portal then restart the VM I am back to the original issue.

When I try to login to the server using RDP, the VM freezes after login as its loading server manager and after sometime the VM agent will go not ready.

If I disconnect the 2 additional disks, the VM unfreezes.

I have also tried simply creating a new disk from the portal and attaching that to the VM.

I then start the VM and can login fine.

I go to disk management and initialise the disk and format it.

after formatting I can see the new drive in file explorer fine.

If I then shutdown the server from the OS then stop the VM from the portal, Then restart the VM from the portal.

I am again back to the same issue when I login, It again freezes at the server manager screen.

while its frozen if I detach the new disk it will unfreeze and allow me to login fully and runs ok.

again these issues only started after the July KB5062553 update

Hello, @Mike Collins ! Following up on my PM, this is what I see so far:

• You originally reported that the July KB5062553 update was applied to your 2025 Azure VM servers which resulted in:
  • Agent is now in "Not Ready" state; agent version is "Unknown"
  • Loss of RDP access
  • VMs are hung in a running state
  • Impact is to exchange server and domain controllers
• Applying KB5064489 does not resolve the issue:
  • Originally could not connect to the servers; no access to the CMD via serial console
  • Subsequent attempt was successful in installing the KB, however logging in via the GUI results in the VM freezing and the agent dying
• Other troubleshooting steps included:
  • Redeploy to move to a different host. RDP briefly enabled until connected and then unresponsive moving forward.
  • Attempted to log in via Bastion, but disconnected as system is not responsive.
  • Resizing did not work
  • Snapshot works, but only without the data disks
  • Attaching data disks results in VM freeze; removing data disks unfreezes VM

This is all correct but the connecting data disks

if connected to an already running vm they can be read and the vm is ok.

if the vm is then restarted, then it will freeze again

Thank you for the additional troubleshooting results, @Mike Collins ! I'm going to include some of the results here so that we have a record of it.

There seems to be an issue with deallocating the VM that is causing it to go into a bad state.

The VM will work normally if you:

• Add 1 attached data disk at a time (not 2, as that will result in the same freezing behavior)
• Then either reboot the server or shut down the server (from the OS) and start the VM from the portal

The VM will not work normally if you:

• Deallocate (stop) the VM and then boot it again

VM (not Azure) on local Hyper-V 2016 does not boot. Tested on multiple separate VMs. Literally no way to do anything (but restore from backup, before it got fully buggered!)