Version 4.11

Troubleshooting

This topic lists errors you might see while migrating VMs.

Troubleshooting VM upgrade during migration

Insufficient space for upgrade

The upgrade process requires a significant amount of free space on the boot disk, used to back up the existing Windows installation. Because of this, an upgrade operation requires boot disk space that can support the upgrade.

  • If there isn't enough space, you'll received an error message such as the following:

    There is not enough free space on partition (C:). A total of {space minimum}
    megabytes (MB) of free disk space is required. Try Disk Cleanup, or moving
    files to an external location such as a CD, DVD, or external hard drive,
    and then restart the installation.
    

    For more about upgrading VMs during migration, see Upgrading Windows Server VMs.

Troubleshooting VM migration

Errors while attaching to storage

When you execute the Run in Cloud or Full Migration operations, Migrate for Compute Engine verifies that the Backend successfully connects to all volumes before continuing.

  • If the timeout limit is exceeded while waiting for all volumes to connect, the verification task fails with the following error message:

    Failed to attach virtual machine disks
    
  • If VMware returns an error code, the code will be added to the error message as follows:

    Failed to connect volume {volume index} of VM {VM name}: {vSphere error}
    

Confirm that the Migrate for Compute Engine backend component has visibility to your VM storage. For example, confirm that network firewall rules permit access.

Errors while booting Linux VMs

When booting your migrating Linux VMs in Compute Engine, Migrate for Compute Engine runs code to adapt the VMs to the Compute Engine environment. If you didn't install the Linux prep package with this software, you might see the following errors.

  • If the package is not installed, booting your VM on Compute Engine can generate the following error:

    Migrate RPM package is not installed on {VM name}
    
  • If the boot fails before the package's installation can be verified, VM boot will fail with the following error:

    Please make sure that Migrate RPM package is installed on {VM name}
    

Install the package to resolve the issue.

Errors while booting Windows VMs

When booting your migrating Windows VMs in Compute Engine, Migrate for Compute Engine runs code to adapt the VMs to the Compute Engine environment. In cases where this process fails, you might see the following errors.

  • If high availability support could not be installed, you can override the error with a runbook setting, setting feature-flags:migrate-storage-ha to FALSE, or by using offline migration.

    "Operation system adaptation process failed to install packages required for
    storage high availability. This can be overridden by setting migrate-storage-ha
    to FALSE, or by running Offline Migration."
    
  • If the migrating VM OS is an supported version of Windows, confirm that it is included in the supported OSes list.

    "Operating system adaptation process discovered an unsupported Windows version
    (Winver={version}). Please refer to Migrate for Compute Engine documentation for 
    supported O/S list."
    
  • If there is insufficient space to adapt the VM, confirm that the boot device has the space you need.

    "Not enough disk space on boot drive. The operating systems adaptation process
    requires at least {minimal_size_gb} GB of free space on boot device."
    

Troubleshooting vCenter SSL certs

vCenter SSL certificates might change as part of routine certificate refresh cycle. When the cert is changed, Migrate for Compute Engine will issue an error stating that vCenter SSL certificate is not trusted.

If this occurs:

  1. Sign in to your Migrate for Compute Engine Manager.

  2. On the Home page, click System Settings.

  3. Select the vCenter Environment tab.

  4. Unregister the vCenter environment. You will be prompted to accept the new certificate.

  5. After the unregister completes, reregister the vCenter.

    It might take seveal minutes to propagate the change.