Jump to contentJump to page navigation: previous page [access key p]/next page [access key n]
ContentsContents
Virtualization Guide
  1. Preface
  2. I Introduction
    1. 1 Virtualization technology
    2. 2 Virtualization scenarios
    3. 3 Introduction to Xen virtualization
    4. 4 Introduction to KVM virtualization
    5. 5 Virtualization tools
    6. 6 Installation of virtualization components
  3. II Managing virtual machines with libvirt
    1. 7 Starting and stopping libvirtd
    2. 8 Preparing the VM Host Server
    3. 9 Guest installation
    4. 10 Basic VM Guest management
    5. 11 Connecting and authorizing
    6. 12 Advanced storage topics
    7. 13 Configuring virtual machines with Virtual Machine Manager
    8. 14 Configuring virtual machines with virsh
    9. 15 Managing virtual machines with Vagrant
    10. 16 Xen to KVM migration guide
  4. III Hypervisor-independent features
    1. 17 Disk cache modes
    2. 18 VM Guest clock settings
    3. 19 libguestfs
    4. 20 QEMU guest agent
    5. 21 Software TPM emulator
  5. IV Managing virtual machines with Xen
    1. 22 Setting up a virtual machine host
    2. 23 Virtual networking
    3. 24 Managing a virtualization environment
    4. 25 Block devices in Xen
    5. 26 Virtualization: configuration options and settings
    6. 27 Administrative tasks
    7. 28 XenStore: configuration database shared between domains
    8. 29 Xen as a high-availability virtualization host
    9. 30 Xen: converting a paravirtual (PV) guest into a fully virtual (FV/HVM) guest
  6. V Managing virtual machines with QEMU
    1. 31 QEMU overview
    2. 32 Setting up a KVM VM Host Server
    3. 33 Guest installation
    4. 34 Running virtual machines with qemu-system-ARCH
    5. 35 Virtual machine administration using QEMU monitor
  7. VI Troubleshooting
    1. 36 Integrated help and package documentation
    2. 37 Gathering system information and logs
  8. Glossary
  9. A Configuring GPU Pass-Through for NVIDIA cards
  10. B GNU licenses
Navigation
Applies to openSUSE Leap 15.3

30 Xen: converting a paravirtual (PV) guest into a fully virtual (FV/HVM) guest Edit source

This chapter explains how to convert a Xen paravirtual machine into a Xen fully virtualized machine.

Procedure 30.1: Guest side

In order to start the guest in FV mode, you need to run the following steps inside the guest.

  1. Prior to converting the guest, apply all pending patches and reboot the guest.

  2. FV machines use the -default kernel. If this kernel is not already installed, install the kernel-default package (while running in PV mode).

  3. PV machines typically use disk names such as vda*. These names must be changed to the FV hd* syntax. This change must be done in the following files:

    • /etc/fstab

    • /boot/grub/menu.lst (SLES 11 only)

    • /boot/grub*/device.map

    • /etc/sysconfig/bootloader

    • /etc/default/grub (SLES 12 and 15 only)

    Note
    Note: Prefer UUIDs

    You should use UUIDs or logical volumes within your /etc/fstab. Using UUIDs simplifies the use of attached network storage, multipathing, and virtualization. To find the UUID of your disk, use the command blkid.

  4. To avoid any error regenerating the initrd with the required modules, you can create a symbolic link from /dev/hda2 to /dev/xvda2 etc. by using the ln:

    ln -sf /dev/xvda2 /dev/hda2
    ln -sf /dev/xvda1 /dev/hda1
    .....
  5. PV and FV machines use different disk and network driver modules. These FV modules must be added to the initrd manually. The expected modules are xen-vbd (for disk) and xen-vnif (for network). These are the only PV drivers for a fully virtualized VM Guest. All other modules, such as ata_piix, ata_generic and libata, should be added automatically.

    Tip
    Tip: Adding modules to the initrd
    • On SLES 11, you can add modules to the INITRD_MODULES line in the /etc/sysconfig/kernel file. For example:

      INITRD_MODULES="xen-vbd xen-vnif"

      Run mkinitrd to build a new initrd containing the modules.

    • On SLES 12 and 15, open or create /etc/dracut.conf.d/10-virt.conf and add the modules with force_drivers by adding a line as in the example below (mind the leading whitespace):

      force_drivers+=" xen-vbd xen-vnif"

      Run dracut -f --kver KERNEL_VERSION-default to build a new initrd (for the default version of the kernel) that contains the required modules.

      Note
      Note: Find your kernel version

      Use the uname -r command to get the current version used on your system.

  6. Before shutting down the guest, set the default boot parameter to the -default kernel using yast bootloader.

  7. Under openSUSE Leap 11, if you have an X server running on your guest, you need to adjust the /etc/X11/xorg.conf file in order to adjust the X driver. Search for fbdev and change to cirrus.

    Section "Device"
              Driver       "cirrus"
              ......
              EndSection
    Note
    Note: openSUSE Leap 12/15 and Xorg

    Under openSUSE Leap 12/15, Xorg will automatically adjust the driver needed to be able to get a working X server.

  8. Shut down the guest.

Procedure 30.2: Host side

The following steps explain the action that you need to perform on the host.

  1. To start the guest in FV mode, the configuration of the VM must be modified to match an FV configuration. Editing the configuration of the VM can easily be done using virsh edit [DOMAIN]. The following changes are recommended:

    • Make sure the machine, the type, and the loader entries in the OS section are changed from xenpv to xenfv. The updated OS section should look similar to:

      <os>
                <type arch='x86_64' machine='xenfv'>hvm</type>
                <loader>/usr/lib/xen/boot/hvmloader</loader>
                <boot dev='hd'/>
      </os>
    • In the OS section, remove anything that is specific to PV guests:

      • <bootloader>pygrub</bootloader>
      • <kernel>/usr/lib/grub2/x86_64-xen/grub.xen</kernel>
      • <cmdline>xen-fbfront.video=4,1024,768</cmdline>
    • In the devices section, add the qemu emulator as:

      <emulator>/usr/lib/xen/bin/qemu-system-i386</emulator>
    • Update the disk configuration so the target device and bus use the FV syntax. This requires replacing the xen disk bus with ide, and the vda target device with hda. The changes should look similar to:

      <target dev='hda' bus='ide'/>
    • Change the bus for the mouse and keyboard from xen to ps2. Also add a new USB tablet device:

      <input type='mouse' bus='ps2'/>
                <input type='keyboard' bus='ps2'/>
      <input type='tablet' bus='usb'/>
    • Change the console target type from xen to serial:

      <console type='pty'>
                <target type='serial' port='0'/>
      </console>
    • Change the video configuration from xen to cirrus, with 8 MB of VRAM:

      <video>
                <model type='cirrus' vram='8192' heads='1' primary='yes'/>
      </video>
    • If desired, add acpi and apic to the features of the VM:

      <features>
                <acpi/>
                <apic/>
      </features>
  2. Start the guest (using virsh or virt-manager). If the guest is running kernel-default (as verified through uname -a), the machine is running in Fully Virtual mode.

Note
Note: guestfs-tools

To script this process, or work on disk images directly, you can use the guestfs-tools suite (see Section 19.3, “Guestfs tools” for more information). Numerous tools exist there to help modify disk images.

Print this page