Skip to content

VM Issues

Common problems with Azure Local virtual machines in the ACX Lab and how to resolve them.

VM Won't Start

Error: "Allocation failed"

Cause: The Azure Local cluster doesn't have enough capacity to start your VM.

Fix:

  1. Try starting the VM again in 15–30 minutes (capacity may free up)
  2. Check if you have other VMs running that you can deallocate
  3. If using a GPU VM, GPU capacity is especially limited — check with the ACX Lab team
  4. Consider using a smaller VM size temporarily

Error: "VM is in a failed state"

Cause: A previous operation on the VM failed and left it in an inconsistent state.

Fix:

  1. In the Azure portal, go to your VM
  2. Click Redeploy + reapply in the left menu
  3. Click Redeploy — this moves the VM to a new host node
  4. If redeployment fails, try Stop (deallocate) then Start
  5. As a last resort, delete and recreate the VM using the provisioning runbooks

VM is Slow or Unresponsive

Check Resource Utilization

  1. In the Azure portal, go to your VM → Monitoring → Metrics
  2. Check CPU percentage, Available memory, and Disk IOPS
  3. If CPU is consistently above 90% or memory is exhausted, consider resizing

Common Causes

Symptom Likely Cause Fix
High CPU, low memory CPU-bound workload Resize to more vCPUs (e.g., D4s_v3 → D8s_v3)
Low CPU, high memory usage Memory-bound workload Resize to more RAM
High disk latency Disk I/O bottleneck Check for runaway processes, consider Premium SSD
General sluggishness Windows Update running Wait for updates to complete, or schedule them off-hours

Resize Your VM

  1. Deallocate the VM (Azure portal → VM → Stop)
  2. Go to Size in the VM settings
  3. Choose from allowed sizes
  4. Click Resize
  5. Start the VM

Can't Connect to VM After It's Running

  1. Verify the VM is in a Running state (not Starting, Stopping, or Deallocated)
  2. Check that the VM's guest OS has finished booting (try waiting 2–3 minutes after status shows Running)
  3. Ensure you're using the correct connection method — see Connecting to Resources
  4. Check the VM's boot diagnostics: VM → Boot diagnostics → Screenshot to see the current screen state

VM Lost Network Connectivity

Cause: The VM's network interface may have been misconfigured or the VNet peering is broken.

Fix:

  1. In the Azure portal, go to your VM → Networking
  2. Verify the NIC is attached and has a private IP
  3. Check Effective network security group rules to ensure traffic isn't being blocked
  4. Try restarting the VM (this re-applies network configuration)
  5. If the issue persists, contact the ACX Lab team — the VNet peering may need to be re-established

Data Loss Warning

!!! danger "No backups by default" Lab VMs do not have automatic backups configured. If your VM is deleted or its disk is lost, your data is gone. Save important work to:

- A Git repository
- Azure Storage (Blob or File Share)
- Your corporate OneDrive

VM Extension Failures

Azure VM extensions (e.g., custom script, monitoring agent) may fail on Azure Local VMs due to Arc agent connectivity.

Fix:

  1. Go to your VM → Extensions + applications
  2. Check the extension status
  3. If an extension shows Failed, try removing and re-adding it
  4. If the Arc agent is disconnected, restart the himds service on the VM:
# On Windows
Restart-Service himds

# On Linux
sudo systemctl restart himds