Migration Hyper-V VM to MCS
Migration of virtual machines between services / platforms on which they can be launched and operated, as a rule, consists of several stages:
Preparing for migration
Before migrating a virtual machine, you should perform preparatory actions aimed at providing the existing virtual server with functionality:
The first step is to download and install the VirtIO package drivers into your existing virtual machine.
In some cases, if automatic installation of drivers is not possible, you may need to manually add them and install drivers from the downloaded VirtIO driver package.
After installing the drivers, you need to install the QEMU Guest Agent. An installer is used that matches the architecture of the existing virtual machine.
Adding drivers to the registry
To correctly identify the drivers, you need to add them to the Windows system registry.
To do this, you should:
- Upload Virtio Registry File to your instance.
- Open Windows Registry Editor and import the downloaded file.
Exporting a virtual machine
There are several ways to export VMs from Hyper-V:
To export a virtual machine to a file, select the desired VM, then export it using the Export function in the virtual machine menu.
In the export window, specify the path to save the exported object.
You can export the required virtual machine using the command
Export-VM -Name <virtual_machine_name> -Path '<full_path>'
Loading VM image into MCS
The resulting * .vhdx file should be loaded into an existing MCS project.
It is recommended to use the Openstack CLI to load the virtual machine image in order to avoid possible errors in processing large files by the web interface. To load the * .vhdx image, use the command:
$ openstack image create --private --container-format bare --disk-format vhdx --property store = s3 --file <file.vhdx> <image_name>
If the instance created from the image must support backup, you must load it with the metadata of the presence of the guest agent:
$ openstack image create --private --container-format bare --disk-format vhdx --file <file.vhdx> --property hw_qemu_guest_agent = yes --property store = s3 --property os_require_quiesce = yes <image_name>